“se pererrt tet r ee eens “ Lawl eae) ns ‘ deed Lott Wott xeestiiwede lemme ah dad tee @ Sahel ot Sane ee A oat hak ot hd, ei eterees 7) te Ce astintt ede ed de ehh fe Paedetaed aad dite Shen wow n-srd ty sig te dade EMS lA AAR ast ae Oe ee aN an seat tsiiearleu adr! BL tne Md df de ede aa id ravnpsiesl dimspanipieieoeceuer cate itbab Tai aus wisdrurarards bed Se te rH dell Penner Leohpssocana e este toaTend wade bdog sav a dels ee P deree eeu v4 wea hin Pete CRT ae wren ie ear ane Leon Chicualtttet boone 149440 A fdut andres het errant 4 SStiedertee 2uive 4 ‘ as Adak set a WT peetee ta 4 Under e-dantde dbew ed unk ta arden deg Ned te Tce rset pie set o ory ae: ie cat sev cafuehae vaiiisa's 1 dels Bret Godke B8 1 ded: Bo ae hanes’) ut ry Wit cods ral dla rey edn aceay Ase eae aia ‘ ee i tribea oo Meahoe eho iy Borin ie Taub a an tate i ededl cee ai branh ease Usk yeni vod ast reed tsbeagen i dtr 4 Tye COMME Nir at den dente Mot ge Porn SE naa ded be bad naa Hiei ower i Seda de bade ba } pW nets sabe aa tp eH aa td aire oad ti Vs Poko dks ea oe all Adc 6 de bead VATE Be Sully phe ee ee baa LR Rat ‘Ws me ay Mead bated Leer See Mereletetee o er + Sehane yori nice see Pied Halifence fatata! " < Meese rine a va ercindend oo dra seeded eed. Soke bate! i Pee dstee ge dal Sededehe Ly Cote tuan male be 5 gededopi delice " Tee fata dele han ihe 08 Fe AH ah antes a ess fet queda tod a at OR BO 8 To) Daas VM be ha 4 ea oe iti esata g a cf Me ha dace Ail at deg weye fe mseiaae Ria escort 58 Aly. 0 eg aid aa N aa Pa Ny honey} ited ai aca tae See idedp aed ohare Blee Dead bee rb PoP aA Oidedee Caled fod THe iet dena iedoferd de) pede ae tided AGE poke asked BaP amped aa Stelios Ley eet Downy m8 deans tated 4 ed tw Spare Naa dea ata’ ender! D4 jared dt band Serenet Hed deO ded de ded ed tN) jaded de HV Mode CP er Pain dte as re Misses ted aati sed Murvedodited tan vats Via: ele WA esto! eer) PT Ee ted 4h as Nhe Fea tbh Beil ledsiiom Ada se bere tore, On Wied PMa dh ke deeded eb hed asd POWER Le ad del AREAL ieled Sse 4 fe eC bedenle renee ter isd atid Tatprusich tdci ORS ey epee er iY: Pediat trem ine My B4diee i dP Saunt “ae bea dedeaa ehncvin by sirk tad vin Shy BAS Asante Pisrnescaeeh tN $4 buse nh eaie Ward 6 a) Pan pei ihe ash 4 Wedd areal a aya Feta tihite bet at dade data Riera st ero ter wiles Pears pe Oe Bae Tan CALCU sida dre anineee 4 adie hie A a dedeth 4 CNW nea Haein vara Meardtdox prveetibedes AC VPs ikdete en obs ide Geded Wg eb wr diirne Fedagh bedenca cy hey ett Sa aided wea Avisepdeaist a a Mei ieee ae yok a ca Had A alae dole ssacued Te etna ES Ansegnp td Ogee Serer Fig Aedtilies & eee eee peewee se met) aed dash ink ree Sq dided de WV hon dosh dot rt Peernn yon iene? i ‘ Cie piacere RATT Sega heal AA "eds hela iim ek kd ie NeW hich bide goa Fh omen ES aire eu trigrt bom VPLS, #8 Moh ainda pt Or (Mi lickrerteten care hy AES tema Den nico eds new oe ksh Oh Me dediwee den Mm invete oc * Sree eee adda al 1 Belted ta Ata A Sa tba dete Cdith aed del die dnlea dedi Wed eurére Dee AM ede bday © Dior ha Ge ue pre? nd. eae iD 4 edundatd oWidedobed tt sided ali tele Ged ENO Ie! Cory rte nts oie ed dh ad Vedidh Goh dubia bed: a Fad ed are SP Gore Ler et aa uae Merc tbe ore ‘ fi eC eeae, Bite Fahd) Node b Dey OME C Eon Serre) Ks debva ol Ria delayed di oro nat Pree ner mn ean pyre) oC ibe dpott ie btm Wout +1 He ye a ged CEG bedi deg ish ot 4 wae ead au ae iovedeiea rite dieicrieraltean ta Dwhebede beitiran wets) H408 fet moet perry we sy roan Pome eieriar: Ir? Ue deat Net Wola 4 Sia dida idea de vnireydsvy8 vs beet Heatimned diatg sa aed dara teanoa ie Aen AE Doh acalas ears stalyised Sioutewaiie Mor teeters nea eras oe Ween te 4 er ieisubesaeden ts EA NH th rhea 0 LL 90d set a eddy vedi t Tdtnede th wa be fein dee ht de neo a cj heed Ra Ora} aa Ye 4), i ans seis Gaze fea > cs oat a ee: i eplibasa Uease seuiarasee ks Pea ante aa etuded anncdies fee 8 De bobs Goh Ded det ded ek at 4 Teh e el: we A ORT Y oh ed Ee “inte: rb Meee ts ane Cay Siteeaae sid ro Meuartrd: beet ded Viet bet its wee 0 di i ie fee felis ba Go bo he dee) eet “t atten eke fh BSH ere are fase An rece caai Sates ial Nanievoloracueted net ey hres mtisall ined thes Revie Jeet anil tuibiae Hea MW it or ions cleo pisneatnncnerd atime erat ett aia eee Saosacmas saad Sree ate nes fecsterast ied ae utatar guns ts vi ra Sonate ue enna evel hate teen a} Pe had Soteg Mea Ee Sat dae eh anda if A Labo anita G iyi Lo NS ape Ha a8 vi sy Sade’ OF oe es Ca fritle Arey ray OF te air eu é ined brene pris ‘\ OR if phi irra tend fake a new itd i fudnepeaanssey tative eer A Pinon Rehety ey Cui ial eae sab Fel Hotheid 8 se Fe rob Cade aide Boer? Dein by aye ad a, re it Soe obe foros Pa Siees he ceeau acest Hicing coh dsenitean hae aka haces vipebe Whnane sah Ws oct loerenoutaheane dosray Ne notre ores ua eas Oy 4s Rel bebe eNaaees * (es PSR esan dea oan Vlabveeetedete ade ealt ard ca dei Eien aaeediae ln bedeh @Sbeiioas e Geibeiar eat sey 14 hte dies ee Pt? ib Nae ode se st re UREA SG a Scstetaetd indeed ods tena et a8 AiR) Seo bed wean # beled Hb a dea ve. ary! seni nh: dodo tee: nH na dvenaiceat ae YSN Me eer Vein tegn a Wey ‘aa! Ah A it Ba, PADI De bod be Saye ae Amd cl vay dol vedi, | owt beauty fe Pe hake Ga de Bae dnd 8 dea ier det ie te Le a Eee reciente ert ena SHig en EA HS Satie k Nebeeieteke ah Pai tance eB 4 ei it fie ee te sier anions dated As Wi ails Cente bn mary rca 2 acne SF teed tags ys eats Dycratnatt hn attire een iran uA a Oe aearlhteay iis ieee ne) errors et ray BLES ed Let We A ded ety aie % oahke co A Beda i aa aie hae peta ioetursel ipcmashuets a A ee Porno Per io ve 19 TR Sree rer ee a SUA gt shed Plait Wide t di Ble hiton baste Pt aie sae by BRDU Dag are eed ghia ee ey 41 L aeeard ey yah de rote) Shs Weds dedi © ne LA 6 aemdrb dey) dealde dda OTR: elias i arde WB Asehs bibs aaeaa dads dolce rine 0 eat dea CO armen Cer err ini ae? OR ees be Suceiw a bre ke dead se Ord raw a dedaayarnn ed Cucece ine 8 Gand ery A Ved) dobar tbe ia ly if ry iu’ ere Wei 61 od ase eh Bee W arakrted diya abo ery WW ew odd A ale a are hid CAS CRU CCCI as ca een rea) Dwerirte bes tutor pear Me ie 4 4 Mom = jai ald he Af A Hise dal cere) 1A orn did dnd bn cL ORY aed Can ieee a in aed aoe hy Woh edean dda ee 96 Aidit Ci ad 6 Beata ayia Ardea Waa ahel Avante Ale anete trast? Die sev ahead sa wes eke Ceri owe Cemiry Sei rdw’ Crore etre) Ns ad 4 we bath Seadoo aia Dard CAs Savi deh ck | aedeay aon here g ana Cores . bento we ehaw thd Cd dep g tedee i ads ard Parr teeny CONT eet here ne IF) SPREE Pidewe eid ged 92 Mo Palde ts oe tb Shela ace ete Wt y rey isnt pescieaee iniedse gees Faia nth saviacsbaare tained ortes caries at criaicd pearerare det eet eat pirates na eae tee a ave tithe alec EUvba hatin, heres ie Matt eh.) ” Core eR ibd i SE te et Coe ere eee 3 aienoreers fete) tiesybag ab drsed he % Cilepestleae eae? reas ese yu tive tS NS lee Oar ee ieee east ay ‘ Wenwoete arnt | ae ales Wedded vice ranltions Had Peaidea Fite panna iM ager fle Ordsurdearaoic ye va oiatacarane Coreen: ie phat sa ea wha vee onlay chipitgn wenabtdioed Ne sisoh chai ee drei dee ribeaveedeaevesc wr dch LF ye shitrai seb ngi rdraee i wee p rtiedl edu dly ea te 4) 14 has ses boy i 5 Pali trai ge tiliiiedrdlardog gins Wed edie Sapa ites spina Sea eeren its ey bd dis beare ae o eae yearacere dre urivare fp crasmncbies ay egi gem pia vet. Seriorirmmtnn at pater ence =e er Orerpee ieee So brands Cavewera ov ees edna dnarats) dal ge dedi apg e sa al Fon seed gn iia hy at ree eee poner RS RRC cee wnaiiiraravtrdne soon mr serait indy bedi tees @yaravone-dededi gp Bert Crm creed m Fai dediw ome oe wea eta Ly I, adr scar dodgeanwsasyichifea @edcds Pha elii hese tot dos eden UALR nae A eg ag dee Biter ior ia SS sh manent Romane * c es ie! ease dinars er tetk iene pera “rasta, greed pie OOOnre irre sri See eerie eter area edewn a Sanainias weeds otc ris stata aritartenyersre dre Gere) toe Hist ane tricia oat Ferien) + EO Fn oo ae Soe ed yang hh eR Nt Wid sac bated FS Conn kore nore Cen THE PHILIPPINE JOURNAL OF SCIENCE EDITED BY PAUL C. FREER, M. D., PH. D. WITH THE COOPERATION OF MERTON L. MILLER, Pu. D.; GEORGE F. RICHMOND, M. S. WeDo MUI PHD An Jee COX CV Prem: RAYMOND F. BACON, Pu. D.; CHARLES S. BANKS, M. S. H. D. GIBBS, B. 8.; R. C. McGREGOR, A. B. PUBLISHED BY THE BUREAU OF SCIENCE OF THE GOVERNMENT OF THE PHILIPPINE ISLANDS A. GENERAL SCIENCE. VOLUME III 1908 WITH 101 PLATES, 48 FIGURES, AND 11 MAps. MANILA BUREAU OF PRINTING 1908 va 78322 a 210249 é "4 ~ y ” Y teers : ie : Pee bs 3 3 F: : 4 | II. CONTENTS. No. 1, February, 1908. . Ferguson, Henry G., Contributions to the Physiography of the Philippine Islands: IJ.—The Batanes Islands.......................-...-- Plates I-IX; figs. 1-4; maps 1-3. Schultze, W., New and Little-Known Lepidoptera of the Philippine SS) EE GES Ph ne ee ea rN Plate I. ETT GOT a Ses eA cose eee Ne Meee TIE UE No. 2, April, 1908. IV. Bacon, Raymond I’., Philippine Terpenes and Essential Oils, T.-..... V. Bacon, Raymond F., Philippine Terpenes and Essential Oils, IT.— Walang yil im os OF eee eee eee eae ee Se lbse ete Cen areseciaet ee VI. Richmond, George F., and Musgrave W. E., The Composition of Horlick’s Malted Milk VII. Editorials Plates IIIf. No. 3, June, 1908. VII. Gibbs. H. D., Methyl Salicylate, I—The Separation of Salicylic Acid from Methyl Salicylate and the Hydrolysis of the Ester...... inves) als TX. Walker, Herbert S., Notes on the Sprouting Coconut, on Copra, aN CEs ONG CG OCOMUL «Ol sees see eee eee ene eee etn aera etn en Ee X. Reibling, W. C., and Salinger, L. A., Portland Cement Testing........ Figs. 1-12. PROT IL OGRe ys este arse ee ets ete kate nak era eas en ak eee A Plates I-III. No. 4, September, 1908. DINE, (Calas, We Clomyneir, — CNS: MINN NE ec et peace seen Plates I-IX. Seip Bean, Robert Bennett. A Theory of Heredity to Explain the AUSADOS Ci WAS WAVED TREK CE oe ccee cence becca da crccens te toa egos ree Secoeem SEE Plates I—-VII; figs. 1—5. XIV. Banks, Charles 8., Biology of Philippine Culicidee Plates I-X. XV. Weise, J.. Description of New Cassidide of the Philippine Islands_ XVI. Schultze, W., Life Histories of Some Philippine Cassididee........ Plates I-VI, XVIT. Horn, Walter. Prothyma Schultzei, a New Species of Philippine Cicindeltace: ... Si seers eeeteine: 9 ornare ea her. Page. 101 JEN XVIII. XXII. XXIII. XXIV. XXV. XXVI- XXVII- XXVIII. XXIX. XXX, XXXT. XXXII. XXXII. XXXIV. MXXY. CONTENTS. McGregor, Richard C., Notes on a Collection of Birds from Si- CULL] OX, jeer tT op ey soy rnc Sees ee cnn ne ne eee . McGregor, Richard C., Some Necessary Changes in the Names of Philippine Birds . McGregor, Richard C., Philippine Ornithological Literature, L... . Mendoza, Maria P., Ramirez, Manuel, and Enriquez, Pio Valencia. An Improved Method of Modeling Especially Adapted for the Central Nervous System; Preparation of Brain Models................ Plates I-III. Waitoria lst: pce: aE BE Ae ed Rice nce No. 5, November, 1908. Cox, Alvin J., Philippine Coals as Fuel... et Plates I—-XIII; figs. 1-2. F Gibbs, H. D., Methyl Salicylate, I1—Solubility in Water at 30°... Gibbs, H. D., The Compounds Which Cause the Red Color in Phenol) sos gs te Gibbs, H. D., and Ageaoili, F., On the Detection and Determination of Coconut: Oil ii ca2. 20s a Re ee eee eee Coxe Allan). aia ori an © Veayise seen sateen eee eee Map 1. Cox, Alvin J., Voleanic Tuff as a Construction and a Cement VV art Or BUS eA a Pe RIS SESS OS ace ee ct No. 6, December, 1908. Bean, Robert Bennett. The Benguet Igorots. A Somatologic Study of the Live Folk of Benguet and Lepanto-Bontoc................ Plates I-VIII; figs. 1-13. Smith, Warren D., A Geologic Reconnaissance of the Island of Mindanao and the Sulu Archipelago. I.—Narrative of the Mxpediti ons Meee eels oe oe ee ee ee ee eee Plates I-XIX; figs. 1-4; maps 1—4. Goodman, Maurice. A Reconnaissance from Davao, Mindanao, Over the Divide of the Sahug River to Butuan, Including a Survey from Davao to Mati. Narrative of the Expedition........... Maps 1-2. Seale, Alvin. The Fishery Resources of the Philippine Islands. Part e—Commencial Wishesy xese 22.05 eects oes see Plates I-X; figs. 1-5. IBook notices: 25.0 SA ee ee ee eee 299 409 413 473 501 533 541 Vou. TIT | FEBRUARY, 1908 _ No 1 _ THE PHILIPPINE JOURNAL OF SCIENCE EDITED BY PAUL) GS BREHR. WE Ds /PHy D, WITH THE COOPERATION OF MERTON L: MILLER, Pu. D.: GEORGE F. RICHMOND, M. S. W. D. SMITH, M. A.; A. J. COX, Pu. D. RAYMOND F. BACON, PH. D.; CHARLES S. BANKS, M. S. R. C. McGREGOR, A. B. - PUBLISHED BY THE BUREAU OF SCIENCE OF THE ' GOVERNMENT OF THE PHILIPPINE ISLANDS A. GENERAL SCIENCE MANILA BUREAU OF PRINTIN 1908 PREVIOUS PUBLICATIONS OF THE BUREAU OF GOVERNMENT LABORATORIES. No. 1, 1902, Biological Laboratory.—Preliminary Report of the Appearance in the Phil- ippine Islands of a Disease Clinically Resembling Glanders. By R. P. Strong, M. D. No. 2, 1902, Chemical Laboratory.—The Preparation of Benzoyl-Acetyl Peroxide and Its Use as an Intestinal Antiseptic in Cholera and Dysentery. Preliminary Notes. By Paul C. Freer, M. D., Ph. D. No. 3, 1903, Biological Laboratory.—A Preliminary Report on Trypanosomiasis of Horses in the Philippine Islands. By W. E. Musgrave, M. D., and Norman E. Williamson. No. 4, 1903, Serum Laboratory.—Preliminary Report on the Study of Rinderpest of Cattle ane, Carabaos in the Philippine Islands. By James W. Jobling, M. D. No. 5, 1903, Biological Laboratory.—Trypanosoma and Trypanosomiasis, with Special pretorencs to Surra in the Philippine Islands. By W. E. Musgrave, M. D., and Moses T. egg. No. 6, 19035.—New or Noteworthy Plants, I. The American Element in the Philippine Flora. By Elmer D. Merrill, Botanist. (Issued January 20, 1904.) No. 7, 1903, Chemical Laboratory.—The Gutta Percha and Rubber of the Philippine Islands. By Penoyer L. Sherman, jr., Ph. D. No. 8, 1903,—A Dictionary of the Plant Names of the Philippine Islands. By Elmer D. Merrill, Botanist. No. 9, 1908, Biological and Serwm Laboratories.—A Report on Hemorrhagic Septicemia in pals At as Philippine Islands. By Paul G. Woolley, M. D., and-J. W. Jobling, M. D. No. 10, 1903, Bi i i i (Due to an Organism Resembling the Koch-Weeks Bacillus). By John R. MecDill, M. D., and Wm. B. Wherry, M. D. No, 11, 1903, Biological Laboratory.—Entomological Division, Bulletin No. 1: Prelimi- nary Bulletin on Insects of the Cacao. (Prepared Especially for the Benefit of Farmers.) By Charles S. Banks, Entomologist: No. 12, 1903, Biological Laboratory—Report on Some Pulmonary Lesions Produced by the Bacillus of Hemorrhagic Septicemia of Carabaos. By Paul G. Woolley, M. D. ; No. 18, 1904, Biological Laboratory.—A Fatal Infection by a Hitherto Undescribed Chromogenic Bacterium: Bacillus Aureus Fetidus. By Maximilian Herzog, M, D. No. 14, 1904:—Serum Laboratory: Texas Pever in the Philippine Islands and the Far East. By J. W. Jobling, M. D., and Paul G. Woolley, M. D. Biological Laboratory: Entomological Division, Bulletin No. 2: The Australian Tick (Boophilus Australis Fuller) in the Philippine Islands. . By Charles S. Banks, Entomologist. No. 15, 1904, Biological and Serum Laboratories——Report on Bacillus Violaceus Ma- nile: A Pathogenic Micro-Organism. By Paul G. Woolley, M. D. No. 16, 1904, Biological Laboratory.—Protective Inoculation Against Asiatic Cholera: An Experimental Study. By Richard P. Strong, M. D. No. 17, 1904.—New or Noteworthy Philippine Plants, Il. By Elmer D. Merrill, Botanist. No. 18, 1904, Biological Laboratory.—l. Amebas: Their Cultivation and Etiologie Sig- nificance. By W. BH. Musgrave, M. D., and Moses T. Clegg. II. The Treatment of Intes- tinal Amebiasis (Amecebic Dysentery) in the Tropics. By W. E. Musgrave, M. D- . No, 19, 1904, Biological Laboratory.—Some Observations on the Biology of the Cholera Spirillum, By W. B. Wherry, M. D. No. 20, 1904.—Biological Laboratory: I. Does Latent or Dormant Plague Exist Where the Disease is Endemic? By Maximilian Herzog, M. D., and Charles B. Hare. Serwm Laboratory: II. Broncho-Pneumonia of Cattle: Its Association with B. Bovisepticus. By Paul G. Woolley, M. D., and Walter Sorrell, D. V. S. III. Pinto (Pato Blanco). By Paul G. Woolley, M. D. Chemical: Laboratory: IV. Notes on Analysis of the Water from the Manila Water Supply. By Charles L. Bliss, M. S. Serwm Laboratory: V. Frambeesia = Its Occurrence in Natives in the Philippine Islands. By Paul G. Woolley, M. D. No. 21, 1904, Biological Laboratory.—Some Questions Relating to the Virulence of Micro-Organisms with Particular Reference to Their Immunizing Powers. By Richard P. Strong, M. D. No. 22, 1904, Bureau of Government Laboratories —1. A Description of the New Build- jngs of the Bureau of Government Laboratories. By Paul C. Freer, M. D., Ph. D. Il. A Catalogue of the Library of the Bureau of Government Laboratories. By Mary Polk, Librarian. No. 28, 1904, Biological Laboratory.— Plague: Bacteriology, Morbid Anatomy, and His- topatholozy (Including a Consideration of Insects as Plague Carriers). By Maximilian . Herzog, M. D. 2 No. 24, 1904, Biological Laboratory.—Glanders: Its Diagnosis and Prevention (Together with a Report on Two Cases of Human Glanders Occurring in Manila and Some Notes on pre ace ey and Polymorphism of Bacterium Mallei). By William B. Wherry, No. 25, 190442—Birds from the Islands of Romblon, Sibuyan, and Cresta de Gallo. By Richard °C. MeGreson: : No. 26, 1904, ‘Biological Laboratory.—tThe Clinical and Pathological Significance of Balantidium Coli. By Richard P. Strong, M. D. No. 27, 1904.—A Review of the Identification of the Species Described in Blanco’s Flora de Filipinas. By Elmer D.' Merrill, Botanist. No. 28, 1904.—1I. The Polypodiacee of the Philippine Islands. II. Edible Philippine Fungi. By Edwin B. Copeland, Ph. D. No. 29, 1904.—1. New or Noteworthy Philippine Plants, III. II. The Source of Manila Elemi. By Elmer D. Merrill, Botanist. No. 30, 1905, Chemical Laboratory.—I. Autocatalytic Decomposition of Silver Oxide. Il. Hydration in Solution. By Gilbert N. Lewis, Ph. D. No. 31, 1905, Biological Laboratory.—l. Notes on a Case of Hematebylunia (Together with Some Observations on the Morphology of the Embryo Nematode, Filaria Nocturna). By William B. Wherry, M. D., and John R. McDill, M. D., Manila, P. I. Il. A Seareh Into the Nitrate and Nitrite Content of Witte’s “Peptone,” ‘with Special Reference to Its Influence on the Demonstration of the Indol and Cholera-Red Reactions. By William B. Wherry, M, D. (Concluded on third page of cover.) a b WA JOT STK 4 ek eae et BASH! CHANNEL Yaw Ls! Morth is Mabsten y Slayant® MAP , labayat i) SHOWING THE LOCATION OF THE otnem 1. BATANES ano BABUYANES ISLANDS ba Ag FROM THE CENSUS OF THE PHILIPPINE ISLANDS Gequect 1905 iiSjaa - Sapnt’ (Local names corrected ) BALIRTANG Me pos CHANNEL Brbuyan Claro 1 hee ad L\ BABUTANES ISLANDS lvpirs Rjolan Te esc el Map No. l. THE PHILIPPINE JOURNAL OF SCIENCE A. GENERAL SCIENCE Vou. U1 FEBRUARY, 1908 No. 1 CONTRIBUTIONS TO THE PHYSIOGRAPHY OF THE PHILIP- PINE ISLANDS: Il. THE BATANES ISLANDS.’ By Henry G. FERGUSON. (From the Division of Mines, Bureau of Science, Manila, P. I.) CONTENTS. Loca Ion. PEOPLE. Hisvory. CLIMATE. GEOGRAPHICAL DESCRIPTION. Sabtan. Batan. Inem. The Siayanes. [bujos. Desquey. [sbayat. SUBMARINE CONFIGURATION. GEOLOGIC PROBLEMS. Origin of the agglomerate. Faulting. Alignment of volcanoes. Correlation with Formosa and the Babuyanes. The first paper of this series, Cebu Island, by Warren D. Smith, was published in This Journal (1906), 1, 1043. 66672 iT 2 : FERGUSON. PHYSIOGRAPHY. Formation of the land. The Sabtan Upland. Period of Uplift and Erosion. Sinking of Ibujos and Desquey. Possible Recent Depression. Marine Erosion. Tidal scour. Coral. Vulcanism. Human response and physiographic conditions. Summary. INTRODUCTION. This paper, embodying results of work during three weeks spent in geologic reconnaissance in the Batanes Islands, is intended to be pre- liminary to an article which will deal more fully with the geology and petrography of this group; hence the problems of structural geology will be in the main reserved for the later discussion and I shall here confine myself principally to the physiography. Before proceeding to a description of this group, I wish to acknowledge my indebtedness to Major-General Leonard Wood, commanding general, Philippines Division, to Commissioner Dean C. Worcester, to Mr. William Edmonds, supervising teacher of the Batanes Islands, for his unfailing hospitality and for much helpful information, and to the native teachers and boys of his schools, particularly Mr. Jose Aguedo, for their freely rendered services as guides during my stay in the islands. LOCATION. The Batanes Islands form the most northern portion of Philippine territory and consist of the islands of Y’Ami, Maysanga,~ Mabudis, Siayan, Isbayat, Inem, Batan, Sabtan, Ibujos and Desquey, of which Isbayat, Batan, Sabtan and Ibujos are inhabited. They lie between lati- tude 20° 16’ and 21° 05’ north (21° 13’ north if the “caistence doubtful” Bashi rocks are included), or approximately the latitude of the southern half of the Hawaiian Islands, and between longitude 121° 49’ and 122° 02’ east. Y’Ami, the most northern island, is about 270 kilometers from Cape Engano, the nearest point of Luzon, 107 kilometers from the Ja- panese island of Little Botel Tobago and 160 kilometers from the southern point of Formosa. It is said that on a very clear day the Formosan moun- tains can be seen from the summit of Mount [raya in Batan Island. The Bashi Channel with a minimum depth of 1,009 fathoms separates the islands from Formosa and the Botel Tobagos to the north, while on the 2Where the local name differs from that given on the United States Coast and Geodetic Survey chart, I have followed the local usage. This chart, based on a British survey made in 1845, is badly confused as to local names, and is unreliable in regard to details of topography. On the maps reproduced in this paper the names are corrected to conform to local usage. PHYSIOGRAPHY OF THE PHILIPPINE ISLANDS: II. 3 south the Balingtang Channel (depth of 95 fathoms without bottom) hes between them and the Babuyanes. The Balingtang Islands, lone rocks rising perpendicularly from the sea, le in the center of the Balingtang Channel and form a connecting link between the Batanes and the Babu- yanes groups. PEOPLE. The Batanes people form a separate race, speaking their own language, or languages, for that of Isbayat is different from the language of the other islands. Professor Scheerer * considers the inhabitants of Batan and Sabtan to be of Malay stock, while those of Isbayat are mixed Ma- layan and Papuan. They are kindly, intelligent, enterprismg and ex- tremely industrious. Throughout the Babuyanes and northern Luzon the Batanes people have the reputation for being excellent workers. The two principal islands, Batan and Sabtan, are overpopulated and the arable land is largely taken up, hence there has been considerable emigra- tion and one finds people from the Batanes scattered Une gHaroe the Babuyanes Islands and Luzon. In the days before the Spanish occupation, the constant warfare be- tween the villages made purposes of defense the first requisite in the choice of a village site, hence the inhabitants lived on the hilltops, going down to work in the fields by day, after the manner of the Pueblo Indians of America. ‘The ruins of these old towns are to be seen on the hills above San Vicente (Batan) and Itbod. Itbod was extremely elaborate, being built more in the form of a single fort than a village. Remains of a large cistern and of storehouses show that the inhabitants were prepared to resist a siege, and ruins of small buildings, apparently watchtowers overlooking the cultivated patches, show the precautions which were taken against surprise. It was here that the natives made their only stand against the Spaniards, being overcome by cannon planted on a neigh- boring hill. With the coming of the Spaniards the hill towns were destroyed and the people forced to move into seacoast villages, the sites of which were as a rule dependent upon the presence of gaps in the coral reefs. HISTORY. The group was discovered by William Dampier in 1687 and named by him the Bashi Islands, after an intoxicating drink brewed from sugar cane (now however termed palic by the islanders). Later in the same year three Dominican missionaries visited the Batanes, but after the death of two of them, the survivor returned to Luzon. In 1724 four Dominicans arrived and stayed for a short time, but it was not until 1791 that the Spanish régime was fully established. From September, *Scheerer, Otto: Mitt. der Deutsch. Gesellschft. f. Natur und Vélkerkunde Ostasiens, Tokyo (1906), 11, Pt. I. 4 FERGUSON. 1897, to December, 1899, the islands were under the control of the revo- lutionary government.* The present name Batanes has been used on Spanish maps since the Spanish occupation, but until a few years ago the original name of Bashi Islands was used on American maps, as it is on the English and German of to-day. Professor Scheerer ® considers two groups; the Bashi Group which consists of Isbayat and the small northern islands, and the Batanes which are composed of Batan, Sabtan, Ibujos, and Desquey. Professor KX6t6, in his articles on the Malayan Archipelago and the dependent isles of Taiwan, makes the same division.® CLIMATE." Rainfall.—The records of the Weather Bureau covering Santo Domingo de Basco for the years 1903-1906 are given in Table I. These show a very heavy annual rainfall, the tenth heaviest recorded and no very marked rainy season such as is found on the west coast of Luzon, where the records of Vigan (Ilocos Sur) show that 92.9 per cent of an average annual rainfall of 2,134.1 millimeters occurs during the rainy season from June to October. The seasons in the Batanes may best be defined as a short, dry season from February to May and a long, rainy one. Extreme differences of monthly rainfall such as those between May, 1905 and 1906, are due to typhoons. In May, 1906, two typhoons passed near the islands causing a precipitation of 153.8 millimeters on the 18th and 19th of the month and 390.8 on the 28th, 29th and 30th. Tasie I.—Monthly rainfall at Santo Domingo de Basco for the years 1903 to 1906, inclusive. | Millimeters of rainfall in— Month. ] ] ] | 1903. | 1904. 1905. 1906. | Mean. | ===! | | | January eae ene | 270.1 | 306.8 | 150. 2 327.6 | 163.7 | Rebruany eee le CEO 66.7 85.5 12.5 48.0 | |;eMarchy Se ea es 51.8 20.0 154.8 119.9} 86.6 NaeAnnyill se notte Cam ieee 141.0 | 10.2 | 89.0 207.5} 111.9 Mely.te seins ee eee 201.1 110-3 21.0 677.0 | 252.4 |. ARNG camo etree | 163.6 262.7 | 151.0 90.1 166.8 (aah eeee acetate | 340.7) 406.5) 267.6] 202.0] 304.2 | AN CUSt aa een newer | 910.7 377.0 207.6 127.4| 405.7 September___-_._-_________ 186.9 168. 6 186.1) 424.1] 241.4 | October ever 151.3 370.8 618.6 | 468.7 | 310.8 121.2 123.0 190.8 | 186.4 | 344.3) 368.5 207.6 361.9 | 324.3 | Otel van eee 3,682.3 | 2,364.8 | 2,034.2 | 3,359.4 | 2,860.1 | “From historical note contributed by William Edmonds to Philippine History, vol. 44, 5 Loc. cit. * Koto, B.: Jour. Coll. Sci. Tokyo (1899), 11, IL, 118 and (1900), 13, I, 46. * Statistics from Monthly Bulletins of the Weather Bureau for 1904, 1905, 1906, and Maso, Rey. M. Saderra, 8. J.: The Rainfall in the Philippines, Manila (1907), Weather Bureau. PHYSIOGRAPHY OF THE PHILIPPINE ISLANDS: II. 5 Typhoons.—Vhe islands he in the track of numerous typhoons which often completely destroy the crops and reduce the inhabitants to the verge of starvation. ‘The houses are all built of stone for protection against these winds and at the times of typhoons, nets are stretched across the roofs and anchored to the ground. Typhoons of the first, third, and fifth groups, amounting to 63 per cent of the total number, affect the Batanes Islands.® Temperature—Table II gives the mean monthly temperatures for Santo Domingo de Basco compared with that of Manila for the years 1904 and 1906. It will be seen while there is very little difference in the mean temperature, the range is much greater in the Batanes than in Manila, being over 6° for the former and less than 4° for the latter. Tasre I1.—Mean monthly temperatures in Santo Domingo de Basco and in Manila for the years 1904 and 1906. 1904, | 1906. witoratlt Santo Do- | Santo Do- mingode | Manila. | mingode}| Manila. Basco. Basco. | DERNOAT BY ere 21.4 25.1 22.7| 25.1 February —_- 21.5 25.4 24,0 26.1 March ____ 23:6 26.2 23.8 26.8 26.0 26:9, 26.5 | 29.2 27.3 27.8 27.8 28.7 27.3 27.2 28.4 28.0 27.0 26.7 28.7 27.8 oT || 26.8 28.8 | 27.3 27.2 26.2 28.0 26.7 26.4 26.3 25.9 26.4 24.0 25.0 | 24.5 25.3 21.8 23.9 | 23 25. 0} 25.1 26.1 26 26.9 GEOGRAPHICAL DESCRIPTION. The geologic structure upon which the topography of the islands is largely dependent naturally brings the Batanes into three groups: 1. The islands consisting in greater part of the older rocks, volcanic agglomerate with basic dikes. .'l'o this group belong the Island of Sabtan and southern part of Batan. ; 2. The younger volcanic group, consisting of Mount Ivaya in Batan, the Island of Inem and the small islands to the north of Isbayat, locally known as the Siayanes. 3. The coral limestone group, Desquey, [bujos and most probably Isbayat. 8 Algue, Rev. Jose, J. S.: The Cyclones of the Far East, Manila (1904), 247. 6 FERGUSON. SABTAN. The Island of Sabtan is the southernmost of the Batanes Group. It is about 10 kilometers long by 4 broad and contains five villages, San Vicente, Santo Tomas and Santa Rosa on the east coast, Santa Inés and San Luis on the western side. From Point Natao southward on the east coast to within half a mile of Santa Rosa, there is no marked relief along the shore, with the exception of some high bluffs of volcanic agglom- erate just south of San Vicente. Coral reefs extend over a large part of the coast, although their development here is not as extensive as on the west coast. South of Santo Tomas there is a long, flat stretch near the shore and here sand dunes reaching to a height of about 100 feet have dammed back the waters from the interior, forming a line of small ponds. The land from the eastern shore rises toward the center of the island in a rather irregular series of marine terraces. The materials forming these terraces are stratified pebbles and sand, for the most part entirely unconsolidated, together with limestone, the latter both limestone con- glomerate and coral limestone. These terraces when viewed from the neighboring island, Batan, seem to be beautifully regular, but on closer examination they are seen to be absolutely without continuity, probably due to the fact that where the protecting capping of limestone is absent, the terrace of pebbles and sand is soon worn away. The highest ledge of limestone occurs at an elevation of about 180 meters. It is a quarter of a mile inland and contains poorly preserved Orbitoides. Beyond the last terrace, the country is very deeply dissected and consists of irregular, sharp ridges of volcanic agglomerate, generally much decomposed. A belt of rolling upland between a half mile and a mile wide extends diagonally across the island from Santa Rosa to Santa Inés. This has an average elevation of about 300 meters; it is bounded on the east by a sharp and very irregular escarpment of volcanic agglomerate. The valleys in this upland are broad and the small streams seem to be at grade, the topography being best described as “gently rolling,” in marked contrast to the sharp feature of the irregular ridges below, in general it is a region of physiographic “old age.” This upland rises gradually to the westward, the pass a half mile northeast of San Luis having an elevation of 400 meters, and it ends in a sharp line of cliffs broken only by occasional steep canyons. These cliffs extend along the whole west coast of the island, being lower (200 meters) towards Natao Point. The material is almost entirely volcanic agglomerate, with occasional beds of stratified sandstone and conglomerate which consist entirely of volcanic material. This rock where it occurs is much faulted, the faults being of small throw, generally less than 10 feet, and also somewhat distorted, showing small dips to the west and northwest. Of course, the folding and faulting are not confined to those parts of the agglomerate formation in which the PHYSIOGRAPHY OF THE PHILIPPINE ISLANDS: IL. 7 sandstone occurs, but it is only where there are bedded deposits that distortion is readily distinguishable. Between Santa Inés and San Luis at an elevation of about 7 meters, there is a raised beach consisting of pebbles and a limestone conglomerate which also contains many volcanic pebbles. This beach is of so recent a date that the streams from the plateau have not yet had time to cut channels through it. From Natao Point, southward to Tangel Point, there extends an unbroken coral reef of considerable width. At the bay just north of Tangel is a flat, triangular stretch of marshy land formed by the ponding back of a small stream by the beach. Here there was a village until recent years, but because of its unhealthy situation it was abandoned and the land given over to cultivation, being almost the only piece of cultivated land on the west coast of Sabtan. The southern portion of Sabtan is extremely rugged. It consists of sharp, irregular ridges of agglomerate ending in steep cliffs. The western part is impassible by land and at the time of my visit the sea was too rough to attempt the trip by boat. On the east coast 1 was able to travel as far south as Point Ajao. Here, steep cliffs of agglomerate, often cut by large dikes of hornblende and augite porphyry (f. n.) jut into the sea. No raised beaches or limestone were seen, but for about 8 meters above sea level the rocks were pitted as if by the borings of marine animals. The ridges of agglomerate seem to run in a general south- easterly direction, meeting the shore en echelon and forming a series of small points. The principal ridge, Ceskid Mountain, ending in Ajao Point, shows a remarkably serrate skyline. Fic. 1. CesKID MOUNTAIN AND AJAO POINT. BATAN. Batan Island is about 20 kilometers long, lies in a northeasterly south- westerly direction and varies from less than 2 kilometers to nearly 6 in width. The topography of the island falls into two distinct parts—first, the extreme northern end, northward from Santo Domingo, which is dependent on Iraya Volcano, and second the southern and by far the larger portion, which shows an independent topography which im many respects is similar to that of the Island of Sabtan. Aside from Mount Ivaya, the principal topographic feature is a range of hills extending 8 FERGUSON. diagonally across the island from Mabatuy Point on the west coast to Desiay Point, the southeastern corner of the island. This ridge rises abruptly from the sea at Mabatuy Point and extends southeast for about 2 kilometers where it reaches its highest elevation of about 460 meters at the twin peaks of Mount Matarem, thence it runs nearly south for a little over a kilometer, then turns southeast again, and where the Ithod gorge cuts through the ridge its direction is nearly east. (See Pls. II and III.) Various spurs run off from the main ridge in all directions, and the topography is extremely confusing. A prominent ridge, with minor sub- sidiary ones, runs eastward from Chaua Point across the island. Another ridge extends eastward from Mount Matarem, the latter forming, with the main southeasterly ridge a large amphitheatral valley drained by the streams flowing through the Ithod gorges. The material of these ridges is all voleanic agglomerate, with occasional outcrops of igneous rock and scattered areas of stratified sandstone and conglomerate. The agglome- rate ridges are sharp and ragged in their upper portions and are cut by streams in deep box cafions. The lower parts and the smaller spurs are covered by a thick mantle of decomposed material, generally in the form of a red clay with partly decomposed volcanic pebbles. The gradual downward creeping of this material on the hillside has given a rough stratification parallel to the contour of the hill. The agglomerate ridges form steep cliffs when they reach the sea as they do at Chaua and Mabatuy Points at the coast northwest from Itbod, and along the east coast of the island south from Eskid Bay (north of Mananioy). Where they are fissured, deep sea-caves are produced. x ME iraya MeMatarem eos 8 ei g {hana Fig. 2. South of San Jose de Ibana there is a series of terraces very similar to those of Sabtan, but rather more regular, reaching a maximum eleva- tion of about 275 meters and consisting of stratified voleanic material, sand, gravel and pebbles. Northern Batan, north of a line drawn across the island from San Carlos de Magatao, shows an entirely different style of topography, the features of the landscape being controlled by the extinct volcano, Mount Iraya. This mountain is a beautifully symmetrical cone, its height as given on the Coast Survey chart being 3,806 feet (1,160 meters). The photograph (Pl. VI) shows that it was-once of greater height, an older and larger cone having been blown away in a former eruption and a new PHYSIOGRAPHY OF THE PHILIPPINE ISLANDS: II. 9 one built up in the old explosion crater, while on the southern side there is a hill of tilted basalt which may be the remains of a still older crater. The present crater is horseshoe shaped, being broken down on the north- ern side where the last lava flow, a stream of basalt, has emerged. Numer- ous, small fissures are seen in the crater and one cuts across its southern wall. These fissures are probably the result of earthquakes. Nearly all the lava flows are basalt, but the earliest seem to be andesite, although no petrographic study of them has as yet been made. The end of the flow of basalt from the present crater is exposed in a sea cliff on the northern shore of the island, and this tells in some detail the story of the last eruption of Mount Iraya. The mountain had been quiescent for a period long enough before the eruption to allow a con- siderable stream valley to be cut through the bedded deposits of volcanic débris which form the cliffs of this neighborhood. The renewal of its activity was marked by considerable explosive force which probably blew away part of the northern side of the present cone and nearly filled the valley with a mass of angular fragments of volcanic material. The latter part of this explosive phase was marked by the presence of a number of basaltic bombs. Finally, a stream of basalt several feet thick flowed down this valley, completely filling it.’ Since then the lava flow has itself been buried under the mass of loose material constantly creeping down the slopes of the mountain. The nearly flat region around Santo Domingo and stretching across the island owes its form to the piedmont wash from Mount Iraya. The hyperbolic curve of the mountain must at one time have been continuous from sea level to summit, but marine erosion has cut off the lower end, leaving sea cliffs varying from 15 to 60 meters in height truncating the piedmont plain and of much greater height where the waves have en- croached upon the actual slope of the mountain in the extreme north- eastern part of the island. (Fig. 3.) A Fig. 3. it | 10 FERGUSON. The cliffs of the Piedmont region are composed of stratified sandstone and conglomerate, poorly consolidated, and it is only the constant action of the surf that permits such loose material to form cliffs. The pebbles in the conglomerate beds are all of volcanic origin, and cross-bedding is common in the sand and gravel layers, but no distinct ripple markings appear. I found several pieces of wood embedded in various parts of the cliffs, but never any marine shells. All these facts go to show that the conglomerate is distinctly of terrestrial origin. However, northward from Santo Domingo to Diojo Point there is a series of cliffs of coral limestone interbedded with a distinctly marine conglomerate, the latter containing a large proportion of limestone pebbles. The relations be- tween this limestone and the conglomerate just described, which forms the cliffs bounding the Piedmont area, is not entirely clear, but if the coral has grown upon the conglomerate, there may have been some slight reworking of the material by the waves. The extremely flat plain of Vergnung on the east coast opposite Santo Domingo seems to have been planed off by marine action, as rounded pebbles and shells are found on its surface. (See Pl. VI.) It is of course possible that the limestone area north of Santo Domingo may haye been an earlier uplift against which the wash deposits from the mountains have been built up. Santo Domingo de Basco, the capital of the subprovince and the largest town in the islands, owes its prominence entirely to its situation. It possesses the only harbor in the islands with a stretch of beach un- obstructed by coral, where boats can be hauled up. It is surrounded by the best farming land and in addition it is the central point for the fertile region around Mount Ivaya. Mt Matarem \\ Rock (Reported 1878) San Vicente Nie José de Ibana 4 See level \ fan burs Ibujes Sabtan Batan Cross Section along line A 6 C0 (Plate 1) 1 Scale 100.000 Fic. 4. Fig. 4 is a cross section through the islands of Ibujos, Sabtan and Batan along the line ABCD of map 2, showing the principal features of the topography. ~~ NEW AND LITTLE-KNOWN LEPIDOPTERA OF THE PHILIPPINE ISLANDS. By W. SCHULTZE. (From the Entomological Section, Biological Laboratory, Bureau of Science, Manila, P. I.) RHOPHALOCERA. NYMPHALIDAS. KLYMNIIN &. ELYMNIAS, Hiibn., Verz. bek. Schmetterl. (1816), 37. Elymnias palmifolia sp. noy. (PI. I, fig. 1). 3, fore wing, dark brown with a decided blue iridescence. A large, subapical, white spot, divided into 3 parts by the veins. An elongated streak between first and second median veins, slight traces of two other elongated submarginal streaks. Hind wing; pale brown near costal and inner margins, darker towards the outer margin. Between the veins, 7 yellowish-white stripes, of which there are 2 between median veins I and II. These stripes run nearly to the base of the wing where they suffuse with the general color, but they are very pronounced towards the margin. Near the outer margin, slight traces of light, wavy spots, shin- ing through from the under side. Marginal series of interrupted, crescent-shaped, white spots in pairs between veins and confluent with cilia. Under side of wings lighter; spots corresponding as to position, but differing in size from those above. Posterior to the subapical spot, a very small white one, the streak between median veins, I and II larger than above. Between submedian and first median veins, a broad, whitish streak having a darker one through its middle and showing traces above. On the hind wing the stripes are shorter than above. On the marginal area many irregular, transverse, wavy spots. Length of wing, 5: 39 millimeters. Length of wing, 2: unknown. Tuguegarao, Cagayan, P. I. Time of capture: July, 1905. (Warren Williamson, collector.) Type g, No. 4604, in Entomological Collection, Bureau of Science, Manila, P. I. 27 28 SCHULTZE. This species is very closely related to Hlymnias malis, described by Semper,’ but as he says “The observation that neither the presence nor the size of a spot, but the arrangement of the individual spots, if present, is very constant in the different species of the genera Elymnias,” I feel safe in describing this species as new. HETEROCERA. SESIID 2. ADIXOA, Hamps., Fauna of Br. Ind. Moths (1892), 1, 198. Adixoa tomentosa sp. nov. (PI. I, figs. 2a male, 2b female, 2c 2d, cocoons with pupal skin.) ?, head dark, violet-brown, front steel-blue, white at the sides; palpi white, sides blackish. Collar steel-blue, bordered in front by yellow. Thorax black with a few brownish scales and a violet sheen; a yellow stripe on the inner margin of tegule. Metathorax yellow: abdomen bluish-violet-black, rear margins of the second, fourth and anal segments yellow, the last slightly lighter. Posterior margin of other segments gray-brown, with a few ochraceous scales. Below, on the first abdominal segment, a white triangular spot; posterior margins white. Anal tuft black, with gray and whitish hairs, below lghter. Legs below whitish. Fore-wing, dark, iridescent, violet-brown with 1 hyaline streak in the cell, 4 beyond it, and 1 below. Discocellular yellowish, below the cell towards the base also somewhat yellowish. Cilia with a few ochraceous scales. Hind-wing hyaline, outer border with the color of the fore-wing ; external cilia the same, those of the inner margin lighter to white, disco- cellular yellowish and brown. 3, palpi yellow, blackish at the sides, front white, yellowish-brown on top. Posterior margin of second, fourth and anal segment yellowish. In the male, the anal tuft below is yellowish-white. Length of wing, 3: 8.5 millimeters. Length of wing, ?: 11 millimeters. Manila, P. I. Time of capture: June, 1905. (W. Schultze, collector.) _ ' Type, No. 3345, in Entomological Collection, Bureau of Science, Manila, P. I. The caterpillars live in a vine, Paederia tomentosa Blume, where they are easily found because of the swellings or nodules which they produce on the stems of the plant. The caterpillar makes a blackish, parchment-lke cocoon compressed at both ends; from three to five cocoons are usually found together in a single nodule. 1 Semper. Reisen im Arch. d. Phil. Die Schmetterl. d. Phil. Inseln. (1892), 5, 63. NEW AND LITTLE-KNOWN LEPIDOPTERA. 29 SYNTOMIDA. CERYX;°* Wllgrn. Wien. Ent. Mon. (1863), 7, 140. Geryx macgregori sp. nov. (Pl. I, fig. 7).- ?, dark brown, antenne white at tips, frons with a dark spot in the center. Tops and sides of head whitish.. Collar with a yellow spot at the sides. Tegule with a yellowish spot at the base. A yellowish, median streak on pro-, meso- and metathorax. Abdomen dark brown with 4 longitudinal series of ochraceous spots on dorsum, ventrum and sides respectively ; last 2 segments without spots and lighter. Fore-wing with 5 hyaline spots, a small, elongated antemedial, 2 medial and 2 post- medial, which have a bluish sheen in certain lights. The upper medial spot is in the cell and the lower postmedial one is divided by the dark vein. Hind-wing with a large hyaline spot from the base below the cell to vein TV. Legs dark brown, cox with yellowish spot. Length of wing, d: unknown. Length of wing, ?: 12 millimeters. Montalban, Rizal, P. I. — Time of capture: February, 1906. (Charles 8S. Banks, collector.) Type ?, No. 5189, in Entomological Collection, Bureau of Science, Manila, P. I. I take pleasure in naming this species after Mr. R. C. McGregor, orni- thologist of this Bureau, who has added a large number of insects to our collection. = EUCHROMIA, Hiibner, Verz. bek. Schmetterl. (1818), 121. Euchromia elegantissima Wller. Eugenies Resa omkr. Jordan (1861), 360. Var. diffusihelvola var. noy. (PI. I, fig. 3). Colors of body and markings of the wings as in the regular form of elegantissima. ‘The general color of the fore-wing is brownish irro- rated with yellowish scales, especially near the outer margin. Cilia dark brown. Length of wing, d: 22.5 millimeters. Length of wing, 2: 22.5 millimeters. Manila, P. I. Time of capture: December, 1903. (W. Schultze, collector.) _ Types d and 2, No. 448, in Entomological Collection, Bureau of Science, Manila, P. I. ARBELIDA. SQUAMICAPILLA gen. noy. Proboscis absent ; palpi very minute and porrect ; antenne of the male short, bipectinated to the tips; legs covered with long hairs, which at the tips are scale-like and on the tibie of the fore and hind legs very prom- inent. Male with a long anal tuft of hair scales. Type: S. arenata. *Hampson: Cat. Lept. Phalaenae, (1898), 1, 40 (Syntomoides ) . 30 SCHULTZE. In the classification I place this new genus after the following: Ar- bela,* Moore, P. Z. S. (1879), 411. Squamicapilla arenata sp. nov. (Pt. I, fig. 4.) 3, palpi pale brownish ; head white; collar in front whitish, at the sides dark brown. Thorax; posterior margin of tegule tufted with dark brown hair-scales, as is also the metathorax. Abdomen ayhitish; tuft of brown scales above the first segment; at the sides of the abdomen tufts of pale, yellowish hairs, the last segments (irrorated) with a few brown scales. Anal tuft very long, with its hair scales ochraceous white, but dark brown at the tips. Fore-wing pale ochraceous white, irrorated with a few brownish scales; traces of an antemedial band. At the base of the wing, below the cell, an irregular, dark brown spot and behind the cell another large, brown, trapezoidal one, sharply pronounced. Cilia with an inter- rupted line of brownish scales. The large scales of the fore-wing are very roughly arranged, giving the wing the appearance of being sprinkled with sand. Hind-wing white; inner margin with white hairs. Cilia broad, with traces of darker scales, as on fore-wing. Under side of the wings whitish. ; Length of wing, 3: 17 millimeters. Length of wing, ?: unknown. Manila, P. I. Time of capture: May, 1905. (Charles S$. Banks, collector.) Type, 3, No. 2812, in Entomological Collection, Bureau of Science, Manila, P. I. ARCTITD Al. ARCTIIN &. PERICALLIA, Hiibn., Verz. bek. Schmetterl. (1827), 182. Pericallia integra Wlk. (PI. I, fig. 5.) ?, head, collar and thorax brownish-black; vertex of head yellowish- white, on top of head a dark brown spot, sides of collar and tegule yellow- ish-white. Abdomen orange above, with a blackish band on each segment, except the 2 posterior, which are black with an orange spot at the sides; abdomen below, dark brown. Fore-wing dark brown, a yellow-white band from base and below the cell nearly straight to vein II, then oblique to apex. Hind-wing orange, costal and outer margin at the apex dark brown, conjoined at middle and at the apex to larger ‘spots; two dark brown spots at the posterior margin between veins I and LV; cilia brown, except an inner margin. Length of wing, 2: 20 millimeters. Length of wing, do: 15 millimeters. ®* Hampson: Fauna of Br. Ind., Moths. (1892), 1, 315. NEW AND LITTLE-KNOWN LEPIDOPTERA. 31 Quingan, Nueva Vizcaya, P. I. Time of capture: April, 1905. (Warren Wilhamson, collector.) Type 2, No. 3986, in Entomological Collection, Bureau of Science, Manila, P. I. The male of this species is well known and Hampson * gives a good description of it as does also Semper.® Only the female is new. LIvTHOSIIN®. DEILEMERA, Hiibn. Verz. bek. Schmetterl. (1818), 178. Deilemera browni sp. nov. (Pl. I, fig. 6.) $, palpi yellowish-white, third joint brown, second brown at the sides only. Head, thorax and abdomen light yellowish, the latter slightly darker. Head with black spots on front and vertex; 2 black spots on the collar, 4 on the tegule and 3 on the thorax. Abdomen above with black segmental bands, laterally with 2 rows of black spots. Fore-wing creamy white. A brown, elongated, triangular area extending from the base along three-fifths of the costa. Another brown area from the base along the inner margin to the lower angle then bending upward, not quite reaching the lower angle of the cell, running along the same and bending at vein II downward towards the base, but meeting vein I shortly before the base. Outer area brown, with its inner edges dentate between the veins. The white, inner area forms roughly the letter T. Hind-wing white, with a brown streak along the costa and a very irregular band along the margins except the base of the inner margin. Below, along the inner margin from the base, a small, brownish streak. Length of wing, 2: 25.5 millimeters. Length of wing, d: unknown. Manila, P. I. Time of capture: ? (Rey. Robert Brown, 8. J., collector.) Type loaned, as No. 219 (Brown Collection), to the Entomological Collection, Bureau of Science, Manila, P. I. I take pleasure in naming this species in honor of its collector, Rev. Robert Brown, 8. J. NOCTUID AH. QUADRIFIN &. POLYDESMA, Boised Faun. Ent. Madag., Lep. (1833), 108. Polydesma opala Pagents. Die Lep. Fauna des Bism. Archipels, (1900), 99. ?, palpi grayish-white with a dark brown streak at the sides and a spot on the last jomt. Head, thorax and abdomen grayish-white; a few darker scales in the middle of thorax; anal tuft reddish-brown. Fore and mid- tarsi and hind tibize and tarsi with dark brown spots. Fore-wing; basal *Hampson: Cat. of the Lept. Phal. (1901), 3, 359. ' Semper, Reisen auf den Phil., Die Schm. d. Phil. Inseln (1892), 6, 486. 32 SCHULZE. area grayish-white, reaching at the inner margin from the base to lower outer angle, then from the angle an oblique irregular line to the middle of costa. A large, irregular triangular suffused dark brown spot at the base and upper angle of the wing along the costa. Outer area of the wing suffused brown, a large white apical and a smaller, indistinct sub- apical spot. Below the apical spot another dark brown one; traces of indistinct dark brown and whitish, submarginal, zigzag lines. Cilia at the apex whitish, then pale brown with darker spots, and again at the inner margin whitish. Hind-wing fuscous, especially near outer margin, with an indistinct postmedial and submarginal band. A small, iridescent streak along the inner margin. Color of the cilia the same as on fore-wing but lighter. Below; fore-wing, suffused fuscous with traces of a post- medial and submarginal band; hind-wing with a dark spot at the end of the cell and the postmedial and submarginal bands more distinct as above. $; collar more ochraceous, thorax with dark, grayish-brown hairs. Fore-wing; the brown basal spot not so dark as in the female, with a whitish marking inside of it and its outer edges continued with a suffused, dark brown, irregular, antemedial band. The white area very much suffused with traces of brownish spots. The white, apical and dark brown, subapical spots prominent. Only the male of this spectes is new. Length of wing, d: 15 millimeters. Length of wing, ?: 15.5 millimeters. Manila, P. I. Time of capture: August, 1904. (Charles S. Banks and Rev. Robert Brown, 8. J., collectors. ) Type 6, No. 5711, in Entomological Collection, Bureau of Science, Manila, P. I. REMEGIA,® Guen. Noct. (1852), iii, 312. Remigia intextilia sp. nov. (Pl. I, fig. 8.) 3, head, thorax and abdomen gray-brown, some dark speckles at the tegulw. General color of the wings light gray-brown; fore-wing with 2 dark spots in the cell and a large prominent one at the end of it. Three very indistinct, dark, antemedial zigzag lines and a medial line from below the cell to the inner margin. A broad, irregular, slate-gray, post- medial band from the apex reaching the inner margin before the angle. An indistinct, white marking at the apex and a submarginal series of dark specks. Hind-wing with distinct, antemedial band nearly straight, suffused, indistinct, medial bands which are combined with the irregular, postmedial band. Some blackish specks at the medial line and a series of submarginal specks as on the fore-wing. The hair fringe at the °‘ Hampson: Fauna of Br. Ind., Moths (1892), 2, 527, Sec. IIT (Remigia) . NEW AND LITPLE-IKNOWN LEPIDOPTERA. 33 inner margin pale. On both wings a fine, pale, marginal line. Under side of the wings, the markings are more distinct; fore-wing with a black spot in the cell and one at the end of it. A distinct medial line and postmedial band. Hind-wing with a dark spot near the base, an ante- medial, distinct medial and a postmedial, zigzag line. The area between medial and postmedial lines is darker, forming a band. Length of wing, 3: 24 millimeters. Length of wing, 2: unknown. Manila, P. I. 2 Time of capture: December, 1905: (Charles S. Banks, collector.) Type g, No. 4808, in Entomological Collection, Bureau of Science. Manila, P. I. : CRUSISETA gen. nov. 3, palpi with the second joint thickened and reaching vertex of head, the third short and blunt; antenne bipectinated, the pectinations short. Thorax clothed with long hairs. Abdomen slender, dorsally with hair tufts to fourth segment, then smoothly scaled and an anal tuft of moderate length. Fore and mid femur and tibia hairy, hind tibia and tarsi with very long tufts of hair. Mid and hind tibie with spines. Fore-wing; costa nearly straight, slightly curved towards apex; apex slightly angled, outer margin slightly curved, on the inner margin near the base a moderate tuft of erect hair scales. Hind-wing with apex slightly rounded, inner margin fringed with long hair. Type: C. basipuncta. In classification I place this new genus after the genus Crithote,* Wik. Jour. Linn. Soe. (1864), 7, 182. Crusiseta basipuncta sp. nov. (PI. I, fig. 9.) Palpi dark brown; head, collar and anterior half of thorax dark brown, changing to fuscous on the metathorax. Abdomen fuscous. Legs, except the tarsi, dark brown. Fore-wing with a straight, antemedial, yellowish line, oblique from the inner margin near the base to subcosta. Base and area along the costa purple-gray. A distinct, dark brown spot near the base and upper angle of the wing. Outer area dark brown, growing lighter towards the outer margin, which is purplish-gray. The tuft of hair scales at the inner margin dark brown. Hind-wing fuscous. Length of wing, d: 16 millimeters. Length of wing, ?: unknown. Manila, P. I. : Time of capture: August, 1905. (Charles 8. Banks, collector.) Type g, No. 3984, in Entomological Collection, Bureau of Science, Manila, P. I. “Hampson: Fauna of Br. Ind., Moths (1894), 2, 541. (a) 34 SCHULLZN. GEOMETRID. BoaRMIN &. MILIONIA, WIlk., Cat. (1854), 2, 364. Milionia pretiosa sp. nov. (PI. I, fig. 10.) ?, palpi and front, pale brown, shot with light blue, the sides of head whitish. Collar, thorax, abdomen and legs pale brown. A triangular, ochraceous spot at base of thorax; collar, tegulae, and coxae bemg especially shot with light blue. Fore-wing light ocherous, the apical half to lower angle of the wing, dark brown. An interrupted medial band from costa reaching the inner margin before the angle. At the base of wing an irregular spot, also dark brown. Hind-wing orange, a large, pale brown spot between costa and vein, which is covered by fore-wing. Another large dark brown spet between veins II and V; at the outer margin beginning on the upper angle a large, irregular spot; three round spots also along the outer margin at veins III, 1V and V; the last one being the largest. There are some traces of pale, brownish spots along the veins next to the inner margin. Length of wing, 2: 25 millimeters. Length of wing, d: unknown. Manila, P. I. Time of capture: July, 1905. (Alice and Fritz Worcester, col- lectors. ) Type ?, No. 3591, in Entomological’ Collection, Bureau of Science, Manila, P. I. PYRALID. SCHOENOBIIN&. SCIRPOPHAGA, Treitschke, Schmetterl. Eur. (1832), 9, 1, 55. Scirpophaga virginia sp. nov. Head, thorax, rear edges of abdomen, anal tuft and wings snow-white. The under side of the fore-wings, in the male sex, except the cilia, is grayish, as are fore-legs, and in both sexes the mid tibize with gray spots. Length of wing, ¢: 7.5 millimeters. Length of wing, 2: 8.5 millimeters. Manila, P. I. Time of capture: September and October, 1905. (G. L. Araneta, col- lector. ) 5 Type ¢, No. 4466, and 9, No. 4351, in Entomological Collection, Bureau of Science, Manila, P. I. This species is closely related to Scirpophaga gilviberbis Zell. NEW AND LITITLE-KNOWN LEPIDOPTERA. 35 PYRALIN A. VITESSA, Moore, Lep., E. I. C. (1858), 299. Vitessa splendida sp. nov. (PI. I, fig. 11.) ’, head yellow, third joint of palpi gray; collar and tegule dark, metallic-gray, bordered by yellow. Thorax yellow, growing lighter to- wards metathorax ; at the middle of thorax two confluent spots of which the posterior one is the smaller. Abdominal segments banded dark gray and white; anal tuft orange-yellow. Legs gray, fore-coxe yellowish, middle cox and middle and hind femora white below. Hind tibiz banded with white. Fore-wing dark, metallic-gray. A yellowish, basal, subtriangular spot with the basal edge excised circularly. The spot extending from costa nearly to posterior margin and having its outer side parallel with the bases of 2 white, subtriangular, antemedial spots, the lower of which is the larger. Two postmedial, white spots, one sub- triangular and subcostal, the other trifid * and subtriangular. The outer third of the wing longitudinally striated with a series of ten nearly parallel, whitish lines. Hind-wing white; outer half and a stripe along the costa dark, violet-gray. Cilia white. Length of wing, 2: 21 millimeters. Length of wing, d: unknown. Maao, Negros Occidental, P. I. Time of capture: November, 1902. (Charles 8S. Banks, collector. ) Type 2; No. 4567, in Entomological Collection, Bureau of Science, Manila, P. I. This species is nearly related to Vitessa suradeva, Moore, Lep., E. I. C. P. 299, Pl. VII, fig. 7. PYRAUSTIN &. PY RAUSTA, Schrank, Fauna Boica (1812), 2, 163. PYRAUSTA, Hampson: Fauna of Br. Ind., Moths, (1896), 4, 429. Pyrausta vastatrix sp. nov. (PI. I, fig. 12.) ?, palpi dark ochraceous, white below. Thorax and abdomen ochra- ceous, the last abdominal segments lighter. Fore-wing, ochraceous yellow with a reddish-ochraceous, excurved, antemedial, and a prominent post- medial zigzag line. A small speck in the cell and another at the discocel- lular, extending towards the postmedial line. The marginal and costal areas also reddish-ochraceous. Hind-wing pale, darker towards outer margin. C: However, Freer * showed some years ago that aldehydes react with sodium as if they contained hydroxyl groups, and Nef * and his students have recently proved that aldehydes, in many cases, do assume the enol form before reacting, so that an equilibrium between enol and keto structure is to be expected in these bodies. It is more difficult to believe in such a structure in the case of benzaldehyde, in which body, if we refer to our ordinary conceptions of organic chemistry, we will need 5 Am. Chem. Jour. (1896), 18, 552. ‘Ann. Chem. (Liebig) (1907), 357, 258. TERPENES AND ESSENTIAL OILS, I. 51 to assume that a minute trace of the methylene derivative is always present. However that may be, it is a fact that all substances capable of assuming the enol and keto types respectively act with the Grignard reagent so far as they have been studied, as if composed of the former, thus, acetoacetie ester, urea, thiourea and the amides, etc., all behave toward alkyl magnesium halides as if they contained hydroxyl groups. EXPERIMENTAL. The action upon magnesium of limonene hydrochloride dissolved in absolute ether, is not vigorous, or even fairly complete unless certain rigid conditions are met, but if these are adhered to the reaction takes place smoothly and very rapidly to practical completion. Many experiments were undertaken to obtain the best conditions. The limonene hydro- chloride must be as pure as possible and in most of my experiments it was twice refractioned in vacuo; the ether also must be very pure and ab- solutely dry. The best results were obtained by the usual procedure of removing all soluble impurities by shaking with small portions of water, then drying the reagent over calcium chloride, distilling from sodium wire and finally keeping the ether over sodium wire in a bottle protected by a tube filled with soda lime. The Grignard body which is produced absorbs oxygen very rapidly, and so in most instances my reactions were conducted in an atmosphere of dry hydrogen. It is desirable to have the magnesium as pure as possible, although satisfactory results may be ob- tained with magnesium which is not strictly so if it is properly treated beforehand. The persistency with which limonene hydrochloride refuses to react unless all conditions are strictly met is shown by one experiment in whrch a mixture of Lhmonene hydrochloride, magnesium and ether was placed at ordinary temperature without reaction and finally heated in a sealed tube to 120° to 130° for twenty hours without any change. The followimg experiments gaye a different result : Hzperiment 1—The limonene was obtained from orange peel oil which was twice refractioned and the terpene then distilled over sodium; its boiling point 30° S 30° qo =0.8350; N D> SS ° = 1.4670; AS S107. One hundred and fifty grams of this product were was 174° to 176° at atmospheric pressure. Specific gravity, dissolved in an equal volume of dry carbon bisulphide and dry hydrogen chloride was passed for two working days into the liquid, which was kept in a freezing mixture. The product was then washed with water and dilute alkali, dried over calcium chloride, the carbon bisulphide distilled and the product fractioned in vacuo. One hundred and ten grams of limonene hydrochloride of a boiling point apo of 105° to 107° at 20 millimeters were obtained. Specific gravity, 3 0.9703 ; 30° 2 SUR Soe es : 7 N54 14770; Ae (ell . The following was the analysis: Found Caleulated for C;)H,;Cl (per cent). (per cent). Cl 20.1 20.3 52 BACON. The Grignard reaction was conducted in a 400 cubic centimeter bromine flask, in an atmosphere of dry hydrogen. One hundred and fifty cubie centimeters of absolute ether and 7 grams of magnesium filings were placed therein and the reaction was begun with a few drops of ethyl bromide, 45 grams of limonene hydrochloride were then added, whereupon the reaction continued vigorously ; after it was completed the product was poured over cracked ice and dilute sulphuric acid carefully added. The ethereal layer was separated, dried over calcium chlor- ide, the ether distilled and the product finally fractioned in vacuo with the following results: Fraction No. 1: B. P. 100° to 102° at 65 millimeters; 25 grams, chlorine free. Fraction No. 2: B. P. 102° to ‘110° at 60 millimeters; 5 grams, containing a trace of chlorine. Fraction No.3: The residue in the distilling flask, 3 grams, contains a small amount of chlorine. Fraction No. 1, was redistilled over sodium at ordinary pressure and then yielded 23 erams of an oil boiling between 174° and 176°, with following constants: Specific gravity, 2°°=0,8957; N20=1.4585, a20° =90.3. : 4° D D Fractions Nos. 2 and 3 were united and fractioned over sodium, they gave 3 grams of an oil passing over between 180° and 210°, containing no chlorine. The low boiling portion, No. 1, which posseses an ether-like odor, was again distilled at ordinary pressure in an atmosphere of carbon dioxide and 8 grams from the middle portion were taken for analysis. The constants of this fraction were: 30° J : = N Dp —!.4565; the molecular refraction was M=45.5 caleulated for C,,H,; | =45.3. (1) 0.2420 gram substance gave 0.7720 gram CO, and 0.2820 gram H.0O. (2) 0.1621 gram substance gave 0.5160 gram CO, and 0.1902 gram HO. Found Calculated (per cent). (per cent). (1) (2) CioHis Cio Aig ) 86.99 86.85 86.96 88.2 H 12.98 13.07 13.04 11.8 The result leaves little doubt but that the hydrocarbon has the formula C,,H,,, but to assure greater certainty it was decided to convert it into the completely reduced substance, C,,H,). This hydrocarbon C,,H,,, is probably the dihydro-limonene: CH, CH, va CH gas H.C CH, and identical with the At para menthene which Semmler® obtained by the ° Ber. d. chem. Ges. (1903); 36, 1035. TERPENES AND ESSENTIAL OILS, I. Do reduction of limonene hydrochloride with sodium and alcohol at a tem- perature not higher than 10°. Semmler’s dihydro-limonene had a boiling point of 173° to 174°; specific gravity, 20°=0.829; Ny =1.463; Pol.= 40, while the same hydrocarbon prepared from phellandrene hydro- chloride by reduction with sodium and amy! alcohol has a boiling point of 171° to 172°; specific gravity, 20°=0.829; Np=14601; Pol.=+-25. Experiment 2.—The limonene was distilled from orange peel oil and had the fe} fe} fo} following constants: Specific gravity, 9 =0.8365: eee, NA 1.4680: ts fe} fe} Yate) the hydrochloride, specific gravity, - =0.9675 ; na =14770; a =66.5°. The analysis gave: Found Calculated for C;9H,;Cl (per cent). (per cent). Cl 20.26 20.3 Fifty grams of this substance were now put through the Grignard reaction with 9 grams of magnesium and 150 cubic centimeters of dry ether in the apparatus used for experiment 1, the reaction being started with methyl iodide and iodine. There resulted 38 grams of crude oil (calculated 39.8 grams), con- taining only 0.61 per cent of chlorine, so that 3 per cent of the total hydrochloride used had not been acted upon by the magnesium. This oil was separated into two fractions by careful distillation in vacuo with the following results: No. 1: B. P. 85° to 86° at 40 millimeters, 29.2 grams containing no chlorine. No. 2: A residue of 7.3 grams containing chlorine. Number 1 possessed the peculiar, ether-like odor of the dihydroterpene and gave ; < Bee Oemel leo OO () = BU ie, the following constants: Specific gravity, oe SSS N5p = 1.4880; A Dy The following was the analysis: 0.1468 gram substance gave 0.4674 gram CO, and 0.1725 gram H.O. Found Calculated for Cy oH, (per cent). (per cent). C 86.81 $6.96 H 13.09 13.04 Twenty-eight grams of the oil C,,H,, were now diluted with an equal volume of carbon bisulphide, dry hydrogen chloride passed into the mixture to saturation and the whole kept in ice and salt for ten hours. The product was purified in the usual manner and yielded 23 grams of an oil boiling between 110° and 115° at 30 millimeters pressure, 8 grams of residue remaining in the distilling flask. fo} fo} The following were the constants: Specific gravity, = 0.931; N ee 1.4624; 4 30° 68°. The analysis gav aoe . The analysis gave: Found Calculated for CyoH, Cl (per cent). (per cent). Cl 20.4 20.3 A gram of this oil when treated with sodium ethylate in alcohol, yielded a liquid of very pleasant odor, probably the corresponding ethoxy-derivative. Twenty grams of the chloride so produced were subjected to the Grignard reaction in an atmosphere of hydrogen, 4 grams_of magnesium and 60 cubic centimeters of ether being used, the reaction being inaugurated with a little methyl iodide, it proceeded vigorously. There were obtained 14 grams of a chlorine-free oil 54 BACON. boiling between 80° and 83° at 40 millimeters pressure. It was redistilled over sodium at ordinary pressure, the final yield being 12 grams of oil of a boiling point 171° to 174°. The product is colorless, with an ethereal odor somewhat ; ; 4 aera bre XU 2 30° 30° resembling that of benzene. Specific gravity, 7 =0.8052, Nap }4459, aa =3.7. The analysis gave the following results: 0.2037 gram substance gave 0.6397 gram CO, and 0.2602 gram H.O. Found Calculated for C) Hao (per cent). (per cent). } $5.64 85.72 H ° 14.31 14.28 This oil, C,,H., dissolves slowly in concentrated sulphuric acid (specific gravity 1.84) without warming or the evolution of sulphur dioxide, giving a slightly reddish solution. The oil is recovered unchanged if the solution is quickly treated with water, more prolonged action of sulphuric acid gives higher boiling products containing sulphur. Concentrated sulphuric acid acts violently on limonene as well as on its reduction product C,,H,s, with marked resin formation and the evolution of sulphur dioxide. Bromine reacts with C,,H., by substitution, with the evolution of hydrogen bromide. Sabatier and Senderson* reduced limonene with hydrogen in the presence of nickel sponge at 250° and thus obtained tetrahydro-limonene, the p-methylsopropyl-cyclohexane, C,,H.,,. CH, CH, CH | This substance has a boiling point of 169° to 170° and a specific gravity fo) at 7 of 0.8132. However, this hydrocarbon was probably not pure, as the authors state, “accompagné d’une petite quantite des produits de deboublemént et para diméthyl et para méthyl éthyl cyclo hexane.” Renard‘ obtained from resin oil a hexahydrocymol boiling between 171° and 173° and having a specific gravity of 0.8116 at 17°. No doubt exists from the above experiments but that I have successively passed from C,,H,, through C,,H,, to C,,H.,. An easy method of obtaining tetra and hexahydrobenzene derivatives from terpenes and their derivatives is therefore at hand. It might be questioned that a compound of limonene hydrochloride with magnesium, obtained after the ° Compt. rend. Acad. d. sc. Par. (1901), 132, 1256. TAnn. Chim. Phys. (1884), (6) 1, 230. TERPENES AND ESSENTIAL OILS, I. HY5) usual manner of the Grignard reaction, is really produced in this in- stance, as reactions undertaken to produce the classes of synthetic prod- ucts usually formed by working with similar magnesium derivatives of other halides have not been successful, but the following quantitative experiments remove all doubt of the existence of such a body. Experiment 3.—Ten grams of limonene hydrochloride were subjected to the action of magnesium after the usual method of Grignard. The ether, containing the soluble magnesium addition product was then filtered in an atmosphere of dry hydrogen in a similar apparatus to that employed by Freer * in his work on sodium acetone. The filtrate was decomposed by means of cracked ice and dilute acid in the usual manner and the magnesium contained in an aliquot portion of the aqueous solution then determined. Found Calculated for CjoHis MgCl (per cent). (per.cent). Mg 12.4 V2.2 It will be noted that a shght excess of magnesium was found and it seems probable that this is due to the few drops of methyl iodide which must be added to start the reaction. The ethereal solution after adding the ice and acidifying gave 5.7 grams of C,,H,,, with the usual properties. Another fact to be remembered in considering the formation of the mag- nesitum addition product is that there is always a considerable evolution of heat when the latter is decomposed by water, a fact which is difficult to explain if it is assumed that the magnesium has simply acted on limonene hydrochloride as a reducing agent. : It is not advisable at the present time to discuss at greater length which of the possible bodies of the empiric formula C,,H,, is formed during this reaction. The second reduction product of the formula C,H.) always, so far, has possessed a slight optical rotation and this fact renders it probable that the latter has not been obtained entirely pure. However, the above experiments prove beyond any doubt that limonene hydrochloride reacts with magnesium after the normal manner of the Grignard reaction. It has been shown that alcohol or other compounds containing hydroxyl react with the products of the Grignard reaction according to the fol- lowing scheme: ROH + R’MgX = ROMeX + WHE. The reactions with water which haye been outlined gave this result. The next step was to study the action of aldehydes upon the product of the interaction of magnesium and limonene hydrochloride in the presence of absolute ether, and benzaldehyde was the first representative of the class selected. Haperiment 4.—Forty grams of limonene hydrochloride, 10 grams of magnesium and 150 cubic centimeters of absolute ether were allowed to react in an atmosphere SAnn. Chem. (Liebig), (1894), 278, 123; 283, 38. 56 BACON. of dry hydrogen in a strong flask fitted with a mechanical stirrer. After the change was complete, 30 grams of benzaldehyde, dissolved in an equal volume of absolute ether, were added, drop by drop. The reaction was violent, taking place with a considerable evolution of heat, while at the same time a nearly solid, yellowish-colored substance separated. The product was now vigorously stirred for one hour, it was then treated in the usual manner, the ethereal layer being well shaken out with acids and alkalies. The alkaline solution on acidifying gave 2.3 grams of benzoic acid. As the neutral solution, after distilling the ether, did not readily solidify, it was fractioned in vacwo with the following result: No. 1: B. P. 85° to 90° at 30 millimeters, 38 grams; No. 2: B. P. 90° to 130° at 20 millimeters, 3 grams; No. 3: B. P. 130° to 180° at 20 millimeters, 7 grams; Tar-like residue, 4 grams. Decomposition appeared to take place above 130° and excepting the first frac- tion, there was no indication of a constant boiling substance. No solid bodies could be obtained from numbers 2, 3 and 4, in ice and salt. Fraction number 1 was dissolved in low boiling petroleum ether and treated with an excess of phenylhydrazine.* Thirty-five grams of benzaldehyde phenyl- hydrazone of a melting point of 154° were thus separated, this quantity cor- responds to 18.9 grams of benzaldehyde. The remainder of fraction number 1 consisted for the greater part of C,,H,s. The higher boiling fractions proved themselves to be an inseparable mixture containing much tar. As the principal products of the reaction were benzaldehyde and C,,H,,, it was thought possible that the benzaldehyde had not acted upon the Grignard addition product at all, but on the contrary that the yellowish, nearly solid substance was produced by the action of benzalde- hyde on magnesium powder in a manner similar to the substances formed by the interaction of sodium and benzaldehyde, as noted by Beckmann and Paul.'® To disprove this assumption the action of benzaldehyde on magnesium was studied. Experiment 5.—Five grams of magnesium powder in absolute ether were ren- dered active by means of iodine, and a solution of 10 grams benzaldehyde in absolute ether was then added. No action took place even after the whole was heated on a reflux condenser for one hour. The following experiment demonstrates that when benzaldehyde acts on the product formed by the action of magnesium on limonene hydro- chloride it does so in such a manner as at once to liberate C,,H,,, or im other words, just as if benzaldehyde were an alcohol in this instance. Eaperiment 6.—Forty grams of limonene hydrochloride, 10 grams of magnesium and 150 cubic centimeters of absolute ether were taken. The apparatus was in principle like that.employed by Freer in his work on sodium acetone. A strong, “Tt has been shown in this laboratory that benzaldehyde can promptly and satisfactorily be separated from its solution in petroleum ether by phenylhydrazine. ” Ann. Chem. (Liebig), (1891), 266, 6. TERPENES AND ESSENTIAL OILS, I. 57 wide-mouthed flask was fitted with a stopper with five holes, the latter carrying respectively: (1) The inlet for dry hydrogen; (2) the stirrer set in a mercury trap, (3) a dropping funnel, (4) the reflux condenser which was fitted with a glass stopcock sealed into it, and which was protected at its upper end by a tube of soda-lime and (5) -a tube running to the bottom of the reaction flask. The latter had sealed onto it a funnel fitted with an asbestos filter and was connected with a filter flask by means of a tube and glass stopcock. The filter flask could be evacuated in the usual manner. After the reaction according to Grignard was complete, 25 grams of benzaldehyde dissolved in absolute ether, were slowly dropped into the flask. The usual, violent reaction with much evolution of heat took place and the yellowish, semisolid substance separated, the whole soon becoming so thick that it was impossible to run the stirrer. After one-half hour, absolute ether to dilute was added in small portions, and by closing the reflux and opening the filter funnel connected with a slight vacuum, the reac- tion product was separated into two portions, one soluble and the other insoluble in absolute ether; of course, care was taken thoroughly to wash the insoluble part with absolute ether. Both the soluble and insoluble portions were now decomposed by means of ice and acid in the usual manner, the etheral solutions resulting were separated, dried and the ether distilled. Soluble portion: The aqueous solution obtained by decomposing the soluble portion contains no magnesiwm. The neutral oil when treated with an excess of phenylhydrazine in petroleum ether gave 4.1 grams of benzaldehyde phenylhydra- zone of a melting point of 153°. The excess of phenylhydrazine was removed by means of dilute sulphuric acid, the petroleum ether distilled and the product fractioned in vacuo. Twenty-five grams of an oil boiling between 90° and 93° at 50 millimeters pressure was thus obtained. This fraction contained no chlorine, and after two distillations over sodium at ordinary pressure had the following ° properties: Boiling point, 175° to 177°, thermometer wholly in the vapor; specific fe} afte yates gravity, a = 0.8250; N - =1.4605; A = =72.7°. These properties demon- strate it to be the same compound C,.H,,, produced by the action of water on the body formed by magnesium and limonene hydrochloride. The residual 7 grams in the distilling flask contained a little chlorine and seemed to consist of the diterpene always obtained as a part of the product of the action of magnesium on limonene hydrochloride. Insoluble portion: Thirty-four grams of benzaldehyde phenylhydrazone of a melting point of 154° after one crystallization from ligroin were obtained by the usual methods from this part. There remained 6.6 grams of a tar-like oil after removing the excess of phenyl- hydrazine and from this neither a erystalline solid nor a constant boiling substance could be separated. Experiments 1, 2, and 3 clearly demonstrate that limonene hydro- chloride forms a compound with magnesium of the usual nature of the iignard addition products, the body being soluble in ether and carrying the theoretical amount of magnesium. Experiment 6 proves that benzal- dehyde, acting on this substance, forms a solid body carrying all the magnesium and that this solid when decomposed with water, principally gives benzaldehyde, at the same time the product soluble in ether and free from magnesium contains the dihydroterpene C,,H,., which has been split off from the Grignard body just as if water had reacted with 58 BACON. this product. Thus the principal reaction of benzaldehyde on dihydro- limonene magnesium chloride would seem to be: C,,H,,MgCl-+-C,H,CHO=C,H,COMgC1+-C,,H,.. (Insoluble in (Soluble in ether.) ether.) The evidence for this assumption is made stronger by the following experiments : Haeperiment 7.—The following quantities were taken: Limonene hydrochloride, 30° 30° 30° 4 35 grams; specific gravity, Hor eo 1105 Ay: =54.6°; N Doses magnesium powder, 6 grams; absolute ether, 100 cubic centimeters. The reaction was carried on throughout as in experiment 6, 23 grams of benzaldehyde free from benzoic acid being used. Soluble portion: This portion contained no magnesium. The total oil was 20.2 grams, from which, distilled in vacuo, the following fractions were obtained : No. 1: B. P. 75° to 78° at 25 millimeters, 17.2 grams. No. 2: Residue 2.6 grams. The usual means of separation gave 4.5 grams of benzaldehyde phenylhydrazone melting at 155° and 156°. The remaining oil, after the removal of the phenyl- hydrazine, had the following properties: Boiling point, 174° to 176°; specific 30° 30° 30° gravity, 49 =0.8248; Nene os AS =37°; it was therefore C,,H,.. > Summary of the soluble portion.—Benzaldehyde 2.3 grams; C,,H,s, 14.7 grams; residue (diterpenes, etc.), 2.5 grams. Insoluble portion: The total oil was 31 grams, containing a trace of chlorine. The product, distilled in vacuo, at 40 millimeters gave the following fractions: No. 1: B. P. 80° to 85° at 15 millimeters, 18 grams. No. 2: B. P. 85° to 215° at 15 millimeters, 7 grams. No. 3: 3.5 grams of residue. There were obtained from the above, 22.8 grams of benzaldehyde phenylhydra- zone melting at 156° and a liquid with the constants: Boiling point, 175° to 176°; BYate) ano oO specific gravity, Sy = 0.8254; N eo =1.4590; A te =36.8. There was also iso- lated 1.1 gram of benzoic acid. Summary of the insoluble portion—Benzaldehyde, 12.3 grams; C,,H,s, 7 grams; benzoic acid 1.1 gram, and 8.5 grams of inseparable, high boiling compounds. The presence of the dihydro-lmonene is proba- bly due to imperfect washing of the insoluble solid by the ether. This is readily understood when the sticky nature of this body is considered. Experiment 8—This experiment was not performed in the elaborate apparatus we had constructed, but instead was carried out in a bromine flask in an atmos- phere of dry hydrogen on a water bath. The following quantities were taken: Limonene hydrochloride, 35 grams; mag- nesium powder, 6 grams. After the magnesium addition product had been formed, 23 grams of benzaldehyde were added. The usual violent reaction accompanied by the separation of the yellowish solid took place, the whole being finally heated with steam for fourteen hours, during which time the ether rapidly evaporated, as an ordinary reflux condenser does not hold that solvent in this climate. The TERPENES AND ESSENTIAL OILS, I. 59 product was covered with fresh ether, decomposed with ice and dilute acid in the usual manner and the solvent distilled, 55 grams of an oil, which was fractioned in vacuo, being obtained. No. 1: B. P. 70° to 100° at 15 millimeters, 28 grams. No. 2: B. P. 100° to 150° at 15 millimeters, 12 grams. No. 3: B. P. 180° to 210° at 15 millimeters (metal bath to 280°), 8 grams. No. 4: Tarry residue, 4.5 grams. ° Number | gave 14 grams of benzaldehyde phenylhydrazone and 19 grams of iS =0.8304; No =1.4640. Five grams of benzoin melt- ing at 137° after recrystallization from ligroin and 1.2 grams of benzoic acid melting at 121° were obtained from Numbers 2 and 3. The residual 14 grams was a tar-like oil, which regenerated considerable quantities of benzaldehyde on being boiled with dilute acids. It is probably in greater part a mixture of polymers of benzaldehyde. C,,Hys, specific gravity, It is seen from the above that benzoin, which should be expected in quantity in the residues of all of these reactions, as it should be formed from a benzaldehyde magnesium compound, is really produced if this compound is heated. Haperiment 9—There were used 50 grams of limonene hydrochloride and 10 grams magnesium, the reaction (which in this instance unfortunately did not reach completion) being conducted in the apparatus with a filter tube described in Experiment 6. A small amount of the ethereal solution was filtered, decom- posed with ice and acid and analyzed, giving 2.6 grams of organic liquid, princi- pally C,,H,,, and 0.307 gram magnesium (calculated Mg=12.2, found=10.3 per cent). The remainder was now also filtered from the unchanged magnesium and from the insoluble portion of the ethereal compound of hydro-limonene magnesium chloride, and the filtrate, which now only contained the soluble portion of the addition product, was then treated with 35 grams of benzaldehyde. The usual evolution of heat took place and the yellow solid already described separated, so that the magnesium addition product produced according to the method of Grignard and separated from all other substances by filtration, is in reality the compound taking part in the reaction, any excess of benzaldehyde, or of magne- sium powder, which is generally present, not taking part therein. The yellow solid produced from this soluble portion was now again filtered in an atmosphere of dry hydrogen and well washed with absolute ether; it was finally transferred to another flask and placed under absolute ether, it is designated below as the insoluble portion. The united etherial solution filtered from this insoluble sub- stance, and the ether used to wash it are termed the soluble portion. The soluble portion: This part was free from magnesium and after removing the ether it consisted of an oil which weighed 45 grams in the crude state and which gave the following fractions when distilled in vacuo: No. 1: B. P. 80° to 90° at 20 millimeters, 33 grams. No. 2: B. P. 90° to 110° at 20 millimeters, 5 grams. No.3: A residue of tar, 6 grams. The above fractions, when treated with the usual reagent gave 35 grams of benzaldehyde phenylhydrazone melting at 155°, representing the excess of benzal- dehyde added to the above solution and 14.8 grams of C,,H,, of a specific gravity fo} at ae of 0.8262 and a refractive index, N ou of 1.4584, the remainder consisted D of unchanged limonene hydrochloride and of diterpenes. 60 BACON. The insoluble portion: One and four-tenths grams of the solid remaining under ether was removed, washed with absolute ether, dried rapidly on a porous plate, weighed to a tenth of a gram and thrown into water. The substance proved to be exceedingly unstable in the air, the yellow powder soon becoming very hot, with the separation of a red oil, a behavior similar to that of sodium acetone, ~ while at the same time a very marked odor of benzaldehyde is developed. The powder dissolved rather slowly when thrown into water so that it was necessary to add a little dilute sulphuric acid. The aqueous solution was extracted twice with small quantities of ether, the solvent allowed to evaporate slowly in the air and a crystalline body, which melted at 121° and which proved to be benzoic acid was separated. A very small amount of an oily residue, having the odor of benzaldehyde remained. Magnesium and chlorine determinations were made on aliquot portions of the aqueous solution with the following result: Calculated (per cent) for H Found | ‘ (per cent). C,H;COMgCl CHSCKO Cl 20.71 21.5 11.2 Meg 12.8 14.7 7.9 fo) These figures, considering the method used and the difficulty of wash- 5 ? fo} a) ing out all free benzaldehyde and the certainty of some oxidation during the transfer, agree very well with those calculated. The remainder of the solid was heated for three hours on a reflux condenser with 18 grams of benzoyl chloride. The product was then treated with alkali and the neutral part, on distillation, gave 4 grams of benzaldehyde, 1 gram of a solid, mostly benzoic acid™ and 5 grams of tarry residue. From the alkali 25 grams of benzoic acid were recovered, it is therefore evident that the solid addition product does not react with benzoyl chloride. Experiment 10.—There were used 10 grams of limonene hydrochloride, 3 grams of magnesium and 40 cubie centimeters of absolute ether. The apparatus was arranged as follows: Flask No. 1 in which the Grignard reaction took place, was fitted with a reflux condenser and a filter tube running to flask No. 2; in this second flask, which also carried a reflux condenser, the filtrate from No. 1 was treated with benzaldehyde, and a second filter tube delivered the filtrate from No. 2 after this reaction, into flask No. 3. All parts were arranged so that they could constantly be kept filled with dry hydrogen. After the Grignard reaction was completed in No. 1, the filtrate which was passed over into No. 2 was treated with benzaldehyde and the solid substance which was formed was well washed with absolute ether, all soluble portions and washings being sucked over into No. 3. On final analysis No. 3 was found to contain no magnesium and only C,,H,;, and the excess of benzaldehyde. The solid in No. 2 was placed under. benzene (dried by distilling over sodium wire) and dry oxygen was run into it for eight hours. The reaction is not very marked, this result possibly being due to the fact that the solid forms dense cakes, difficult to penetrate. However, a gradual reddening took place so that the whole, when the current of oxygen was finally shut off had assumed a deep red color. There resulted 1.4 grams of benzoic acid melting at 121°; 0.9 grams benzaldehyde proved by transference into the phenylhydrazone melting at 155° and 1.2 grams of tar. Magnesium and chlorine determinations were made and when calculated on the total substance finally obtained, gave the numbers 11.8 per cent for magnesium and 20.4 per cent for chlorine. Calculated for C,H,;COMgCl, magnesium 14.7 chlorine 21.5 per cent. “ Benzaldehyde oxidizes very rapidly in this hot, moist climate. TERPENES AND ESSENTIAL OILS, I. 61 Experiment 11.—There were employeed 20 grams of limonene hydrochloride, 100 cubic centimeters of absolute ether and 10 grams of magnesium prepared with iodine according to the method of Von Baeyer.” This means of rendering the magnesium active was finally found to be the most satisfactory, as with the metal so prepared the Grignard reaction starts immediately without the aid of any other catalyzer and continues to completion with great vigour. The reaction was carried out in the apparatus used for Experiment 10, 24 grams of benzaldehyde (free from benzoic acid) being added to flask No. 2. The contents of all flasks were worked up in the usual manner. Flask No. 1 gave 5.2 grams of an oil, principally C,,H,s, representing the por- tion of C,,H,;MgCl which was insoluble in the quantity of absolute ether used. The precipitate with benzaldehyde formed in flask No. 2 was very thoroughly washed with absolute ether and immediately worked up; the total oil being 10.6 grams. The following analytical data were obtained: 15.56 grams substance (calculated) gave 1.93 grams magnesium and 3.03 grams 3 chlorine. Found Calculated (per cent). (per cent). Mg 12.4 14.7 Cl 19.48 21.5 From the 10.6 grams of total oil there were isolated 0.2 gram benzoic acid, 9.1 grams of an oil boiling between 177° and 182°, the latter was converted into 16.5 grams benzaldehyde phenylhydrazone melting at 154°, and 1 gram of a tarry residue remained in the distilling flask. Flask No. 3 contained only a trace of magnesium and yielded a total of 21 grams of oil boiling between 170° and 183°; from the latter a quantity of benzaldehyde phenyldrazone corresponding to 12.96 grams benzaldehyde was isolated. The remaining liquid which in amount was 10.8 grams. after removing the excess of phenylhydrazine, was the dihydro-terpene C,,H,,. The calculated amount of pure C,,H,, which would be set free from C,,H,;MgCl by 9 grams of benzaldehyde (formed by decomposing this magnesium compound in flask No. 2) is 12.4 grams, if the course of the reaction is that which has been assumed in this paper. This agreement is a close one when the inevitable losses in frac- tionation are considered. In this experiment, therefore, the reaction has been shown to proceed quantitatively according to our assumptions. The combination of magnesium chloride and benzaldehyde which has been so fully discussed, when freshly prepared is exceedingly unstable in the air, oxidizing upon exposure with a marked evolution of heat and with reddening. When left standing under absolute ether it is fairly stable, especially if present in a thick layer, so that if such a thick layer is allowed to stand over night in the open air but little benzoic acid is obtained, the benzaldehyde on acidifying being in greater part recovered as such. Oxidation is rapid if the solid is suspended in ether and a stream of air or of oxygen is passed in, but under these circumstances peroxide formation takes place, and for this reason benzene was chosen instead of ether. It is entirely premature at the present time to enter into a discussion of Y Ber. d. Chem. Ges. (1905), 38, 2759. 62 BACON. the constitution of this addition product. The simplest assumption would C,H,C—O—MgCl assign to 1 a formula C,H,COMgCl. The structure | C,H,C—O—MeCl does not seem probable, for neither benzoin nor benzil could be found among the substances obtained by decomposing the compound with water, excepting in one experiment (number 8) where heat was used for a long time and where this result was to be expected. However, it is certain that when benzaldehyde acts on the etherial compound of hydro- limonene magnesium chloride, the former reacts as it would do if it contained a hydroxyl group. As in this instance benzaldehyde reacts as if it were an alcohol, it is not surprising to find that acetone does the same, as acetone often assumes the enol form during reactions as, for instance in the production of sodium acetone and during its con- densations, many of which are best explainable on the assumption of a compound of the structure CH,.COH:CH,. It was nevertheless deemed advisable to undertake an experiment with acetone as the reagent. Experiment 12.—There were used limonene hydrochloride, 35 grams; magne- sium, 10 grams, and absolute ether, 100 cubie centimeters. The Grignard reac- tion was allowed to take place in a bromine flask, after it was complete, 13 grams of carefully purified and dried acetone were slowly added. A con- siderable evolution of heat, as is the case with benzaldehyde, was observed and the separation of a solid which at first was red, then yellow and finally white took place. After the reaction has ceased, the product was treated as was the case when benzaldehyde was the reagent. The neutral oils were distilled in vacuo and gave the following fractions: No. 1: B. P. 73° to 80° at 15 millimeters, 20 grams. No. 2: B. P. 80° to 110° at 15 millimeters, 7 grams. No. 3: Residue, 2 grams. Number 1 was identified as CyoHis as it had the following constants: Boiling : eas ZAe . : OO gen 30 cae : point, 174° to 176°; specific gravity, Ig Es Ne oe: Number 2 contained a considerable amount of chlorine and the oil was doubt- less a mixture of unchanged limonene hydrochloride and diterpenes. The method of Deniges ® showed 12.6 grams of acetone to be present in the aqueous solution. This result is parallel with the one obtained with benzaldehyde and therefore, in this reaction, also, acetone assumes the réle of propen— 1-ol-2. These experiences with limonene hydrochloride were of sufficient in- terest to warrant a study of the action of benzaldehyde on other magne- sium organic halides, for it might be true that in each instance a certain proportion of the reduced hydrocarbon might be produced, owing to the small percentage of the enol form present in the aldehyde; with this end in view a study of the reaction between benzyl magnesium chloride 18 Compt. rend. d. Acad. sci. Par. (1898), 127, 963; Ann. Chim. Phys. (1899), (6), 18, 400. Bull Soc. Chim. (1899), (3) 21, 241. TERPENES AND ESSENTIAL OILS, I. 63 and benzaldehyde was undertaken. Thus Grignard™ states that benzyl magnesium halides do not react in the normal manner with aldehydes, the principal product of the reaction being dibenzyl, and Hell*® and his students, depending upon the conditions, obtained both the carbinol and stilben in this reaction. Several experiments under varying conditions were performed and phenyl-benzyl carbinol was always obtained in large quantities, but in no case could toluol be detected. Phenyl-benzyl carbinol is readily ob- tained in the pure state by distillation of the reaction product in vacuo. It boils from 167° to 170° at 10 millimeters pressure and immediately solidifies in the receiver, the room temperature being 30° to 33°. Hell speaks of the difficulty he experienced in obtaining the carbinol in a solid form when working in the summer. Crystallized twice from li- groin (in which solvent when it is boiling it is quite soluble, whereas it is almost insoluble in the cold), it melts at 67° to 68°. The melting points given for phenyl-benzyl carbinol in the older literature are too low.*® With concentrated sulphuric acid it gives a white tar, just as benz- hydrol gives a red tar with the same reagent. As the method of preparing benzhydrol in quantity has in the past been tedious to apply, and as the Grignard reaction was under considera- tion, it was decided to ascertain if it might be available for this purpose. Using chlor- or brom-benzol, magnesium and ether, and treating the reaction product with benzaldehyde, the results leave nothing to be de- sired as to yield, ease of manipulation and time consumed, so that this method will undoubtedly replace the longer ones formerly used in pre- paring this compound. SUMMARY. Limonene hydrochloride reacts with magnesium to form a hydro- limonene magnesium chloride, soluble in absolute ether, the union taking place normally according to the type of reactions discovered by Grignard. This addition product when decomposed by water gives a dihydro-terpene Cio His. A method is developed by means of the Grignard reaction of passing from terpenes and their derivatives to di- and tetrahydro-terpenes and 4 Ann. de VUniversité de Lyon (1901), N. S. 6, 1-116; Chem. Centrbl. (1901), Il, 623. & Ber. d. Chem. Ges. (1904), 37, 453, 225, 1429. * Limpricht and Schwanert: Ann. chem. (Liebig), (1870), 155, 62. Gold- berg: Ibid. (1874), 174, 332. WKnovenagel and Arndts: Ber. d. chem. Ges. (1902), 35, 1987. Sudborough: J. chem. Soc. London (1895), 67, 605. Beilstein II, 1079 gives M. P. 42° probably a misprint for 62°. The error has been copied into Richter’s Lexikon der Kohlenstoffverbindungen. WNef: Ann. Chem. (iebig), (1897), 298, 202. Bacon: Am. Chem. Jour. (1905), 33, 68. 64 BACON. their derivatives, and thus a very simple and accessible means of prepar- ing tetra- and hexahydro-benzene derivatives is at hand. It has been proved by quantitative experiments that with hydro- limonene magnesium chloride, benzaldehyde acts as if it contained a hy- droxyl group. The solid product of the action of benzaldehyde upon hydro-limonene magnesium chloride gives analytical data which point to the formula C,H,COMgCl. When decomposed with dilute acids this compound re- generates nearly quantitative amounts of benzaldehyde, and only when it had been heated for a long time are products lke benzoin obtained, which would indicate a double molecule. The compound C,H,COMgC! is, when freshly prepared, very unstable in the air, its behavior being much like that of sodium acetone. Acetone, like benzaldehyde, also reacts as if it contained a hydroxyl group. No analogous reaction was obtained from benzyl magnesium chloride and benzaldehyde. Work with the Grignard reaction in the field of the terpenes will be continued. PHILIPPINE TERPENES AND ESSENTIAL OILS, II. YLANG-YLANG OIL. By Raymonp F. Bacon. (From the Chemical Division, Bwreaw of Science, Manila, P. I.) INTRODUCTION. The ylang-ylang oil industry is the most important and in fact at present practically the only perfume-oil industry in the Philippines. Like Manila hemp, the ylang-ylang (flower of flowers) is peculiarly a product of the Philippines, as the oil distilled in other tropical countries, prepared from the same tree, is not ranked in the same class, as regards quality, with the product of this Archipelago, but is sold as “cananga” oil. Ylang-ylang oil is obtained by steam distillation from the’ flowers of Canangium odoratum Baill. (Cananga odorata Hook. f. et Th.). Some idea of the magnitude of the industry may be obtained from the following extracts from the reports of the Philippine custom-house showing the amounts of this oil exported during the fiscal years named: TABLE I.—Eaxport of ylang-ylang oil from 1900 to 1907, inclusive. | Year. Kilos. \— | = 11, 847 | 17, 826 | Figures available for | Year. | Kilos. | [ Balers fiat | SL RB 5E = eens eee eae | 1,708 S86 Ses pet nee mee nae | 1, 487 18672 2ei eestor eens | 1,181 A SG8 Ee ease eMC eee ae . 899 || 1.88 9 Ses ee es 1, 080 | 1These figures represent gross weights of the packages and hence do not rep- resent the actual amounts of oil exported, on which no figures are available. To obtain the actual weights of ylang-ylang oil exported the figures given should probably be divided by ten. 65 66 BACON. It is thus evident that the industry is rapidly increasing in mag- nitude, as perfumers are now using this fine oil in an increasing number of products. As the industry is peculiarly one belonging to the Philip- pines, this laboratory has undertaken studies on the oil and the best means of its distillation, with the view of improving the methods of manufacture and of the quality of the oil. The results thus far ob- tained are recorded in this paper. GENERAL TRADE CONDITIONS. The general trade conditions in the ylang-ylang oil industry are not particularly promising for a person with limited capital who desires to engage in this business. Necessarily, the price of the oil depends upon its quality, but the establishment of a brand is of very great importance. Oils of long established brand command a higher price than unbranded ones, or than oils of newer brands, although the latter may be fully as good in every respect. Many European houses buy only through Manila firms with whom they have contracts and will not purchase oil from others, no matter what the quality may be. Nevertheless, there is a large open market, especially with French houses, for first-class oils, and oils of the very best quality can always be sold. The price naturally varies according to the supply and demand, but the very best oils may be counted upon to bring about 200 pesos, Philippine currency, per kilo? and those of established brands bring even higher prices. ‘The conditions in regard to second-grade oils are not by any means as favorable.* The demand for these oils is small, probably because of competition with cananga oil and with artificial ylang-ylang oil. Most of the Manila distillers manufacture both first and second grade oils and because of their trade connections they are usually able to dispose of stocks of the latter. The distillers in the provinces who, because of ignorance of the best methods of distillation and of poor apparatus, usually manufacture only second-grade oils, often find great difficulty in selling their product. These provincial distillers are generally anxious to realize quickly on their stocks, which they sell to the Manila firms for the best price offered. The price paid in Manila for such pro- vincial oils is about 30 to 70 pesos, Philippine currency, per ilo. There is often so little demand for second quality oils that they can not be sold in Manila for any price, and the Huropean market for this grade is frequently so inactive that distillers may have such oils on their hands for as long a period as two years before disposing of them. *One peso, Philippine currency, is equal to one-half a dollar, United States currency. “Interesting in connection with the present prices of ylang-ylang oil is Gals article on this substance (Compt. rend. acad. d. Se. Par. (1873), 76, 1482, in which it is stated that the price of the oil at that time was 2,500 franes the kilo. He calls the oil essence of alan-gilan or hilan-hilan from Unona odoratissima, a synonym of Canangium odoratum. YLANG-YLANG OIL. 67 In general, the distillers do not own their own groves of ylang-ylang trees, and the market for the flowers in the region around Manila is in a very unsatisfactory condition for the distiller. The large number of distilleries in Manila causes keen competition for the flowers; as a result the quality sold is very poor and the price is high. One of the largest firms in this business states that the flowers are of a much poorer grade during the past few years than they were several years ago. The flowers are usually picked in the night and are collected in small lots early in the morning by native brokers who deliver them at the dis- tilleries. The natives make a practice of wetting the flowers with as much water as they will absorb and there will often be leaves, twigs, and other substances mixed in with them to add weight. The distillers hardly dare refuse such materials, although the quality is poor, for fear their supply may be altogether cut off. I should estimate that over three-fourths of the flowers brought to the distillers in Manila are un- ripe and. green, although the ripe, yellow product gives a larger yield of much superior oil. The price of flowers in Manila varies from 20 to 40 centavos, Philippine currency, per kilo, the average probably being as high as 30 centavos.* Many people in the Islands have an idea that the ylang-ylang distillers make a tremendous profit. From my observa- tion, of the business I can not believe this opinion to be well founded. It probably requires on an ayerage 350 kilos of flowers to produce 1 alo of first-class oil and this amount will also probably give an ad- ditional three-quarters of a kilo of second-class oil. Thus the flowers for 1 kilo of first-class oil will probably cost 115 pesos, Philippine currency, and after the cost of steam and water used, of skilled swpervi- sion, interest and depreciation of the plant are added only a legitimate profit is left. The grower seems to be the one who makes the large profits in this industry, as I have heard of trees being sublet by the year for 2 pesos per month, the renter expecting to make a_ profit from the flowers which he can pick from the tree. This would mean a production of at least 80 kilos of flowers from one tree during a year. We have no figures, nor have we been. able to obtain any reliable data -on the point of the yield of flowers from one tree, but there is no doubt but that large trees bear very luxuriantly. In Manila, the best flow- ers are usually obtained in May and June, but the season just passed ‘In Piesse’s Art of Perfumery, London, (1891), 134, is found an interesting error in the statement that ylang-ylang flowers are adulterated with flowers of champaca (Michelia champaca Linn.) to cheapen the quality of the oil. As a matter of fact champaca flowers sell in Manila for as high as 1 peso, Philippine currency, per kilo with a good demand as compared to ylang-ylang flowers at about 30 centavos, Philippine currency, per kilo, and oil of champaca is much more ex- pensive than oil of ylang-ylang. An idea prevails in Manila that the distillers make much money by sorting out the champaca ‘flowers from the ylang-ylang flowers, as in some localities both classes of trees grow in the same grove. This idea is also erroneous. 68 BACON. (1907) was late, so that distillation was not begun on any large scale until August (1907) and it then extended into February (1908). Large numbers of ylang-ylang trees grow in many of the provinces; in the Camarines, Mindoro, and Albay there are stills, and in Bohol there are many trees, but as yet no stills. he trees are also probably found in quantities in many of the other provinces, where oil is not yet distilled. The impression is very general in Manila that the provincial flowers are inferior and will not make good oil. This opinion is no doubt largely due to the lower quality of provincial oils caused by poor distillation. There is every reason to believe that the flowers are just as good in the provinces as in the region around Manila, especially in those regions where the trees are cultivated and raised on a large scale. Some firms in Manila distill oil in the provinces, and the prices which they obtain are just as high as those derived from their Manila product. Moreover, the provincial distiller has two very decided advantages over his Manila competitors in that the price which he pays for flowers is lower (12 to 20 centavos, Philippine currency, per kilo), and in that he can refuse to accept poor flowers, as the competition is not so keen. The next ad- vance for the industry would seem to be the installation of first-class apparatus and the introduction of correct distillation methods in the provinces. METHODS OF DISTILLATION. Much mystery surrounds the distillation of the oil of ylang-ylang in Manila. The manufacturers are supposed to have valuable trade secrets, so that no one is allowed to visit the distilleries of many of them. How- ever, I have been inside of some of the Manila distilleries where no such restrictions exist and have also distilled the first quality of ylang-ylang oil in this laboratory. I do not wish to violate any confidence imposed in me by the manufacturers who have allowed me to visit their plants and have told me of their methods of distillation, but I do not believe there are any trade secrets; by this I do not mean to imply that any one can distill first quality ylang-ylang oil, but with the proper apparatus an operator who thoroughly understands the distillation of essential oils in general will soon find out the special small points in the distillation of ylang-ylang oil. The important pomts where many err, and this is especially true of the provincial distillers, is in the wrong choice of frac- tions, in burning the flowers and in obtaining too much resin in the oil. The oil must be distilled slowly, with clean steam, the flowers being so placed in the stills as to avoid their being cut into channels by the steam. The quantity of the oil taken is only a fraction of the total amount in the flowers. Disregard of this factor is one of the most grievous errors of the provincial distillers, for, on the contrary, they are usually too anxious to obtain a large yield of oil, and therefore they will YLANG-YLANG OIL. 69 often distill 1 kilo from 150 to 200 kilos of flowers. The quantity of the latter to be taken to produce 1 kilo of oil naturally varies with their quality, but in general the amount should be 300 to 500 kilos, probably averaging about 400. After the first quality oil has been distilled, then a varying quantity of the second grade, up to a volume equal to that of the first, may be obtained from the same lot of flowers; after this operation the still and condensers must be thoroughly cleaned and steamed out to prevent contamination of the next distillation of first- quality oil with the remains of the second quality adhering to the ap- paratus. The distiller usually judges of the time to change the receptacle from that used for first quality to that employed for the second, by taking note of the odor of the distillate. The oil is received in some type of Florence flask, usually two or more of these are connected in series and the condensed water is used in future distillations. The whole apparatus is best lined with block tin, although some distillers have found. nickel to be more satisfactory. The oil, after separating from the water, is clarified and as it is sensitive to light and air, it should be placed into dark colored bottles as soon as possible; these should be filled to the neck, well stoppered and then paraffined to keep out all air. In the ideal apparatus the receivers should be so constructed that very little ight and air has access to the oil. The possibilities of vacuum distillation with steam to obtain as large a yield as possible of the fragant lower boiling esters and alcohols and as little as possible of resins and sesquiterpenes, has suggested itself. Some experiments along these lines were under- taken, but the apparatus available was not satisfactory for the purpose owing to losses in the condensers. ‘The maceration of the flowers to allow the oil to escape more easily also suggests itself as a possible improvement in distillation methods. Experiments along these lines will be undertaken at some future time when a nev still, adapted to the purpose, has been purchased. THE ANALYSIS OF YLANG-YLANG OIL. Ylang-ylang oil does not owe its fragrance to any one substance, but contains a great number of odoriferous compounds, hence it is not possible to value it by certain analytical determinations as is the case with many essential oils. At the present time ylang-ylang oil is largely bought and sold on the judgment of the dealers, the determining factor being the odor, and muclr prejudice as well as uncertainty of valuation exists. It is highly desirable to have other means of determining the value of the oil. This would be especially advantageous for those Manila houses who purchase oils other than those of their own manufacture, as ex- perience has shown to these firms that the judgment of the purchaser in Manila does not always agree with that of the one in Europe. It is obvious that it is not possible in the present state of our knowledge of ylang-ylang oil to judge of its quality from the analysis alone, but I 70 BACON. believe the following results will show that the ordinary analytical con- stants are of very great assistance, as only the records of isolated constants obtained on oils largely of unknown origin are available in the litera- ture. Owing to this lack I have determined the simple constants on a number of oils of known origin. Most of the latter were from one dis- tillery, the process of distillation being watched and the samples col- lected by myself. In this manner I was able to obtain a few analytical constants on oils known to be of first or second grade; the results recorded present many regularities and are so promising that it is hoped, as more material becomes available, to discover other constants, so that in time the purchase and sale of ylang-ylang oil may be placed on an exact analytical basis. In this first series the values determined were: Specific gravity BYate) ~ (pyknometer) ; optical rotation at 30°; refractive index, NS and ester number, the latter by the usual method, using 1 gram of oil. The results are tabulated as follows: at TasLe I].—Tabulation of the constants of first-grade ylang-ylang oils. Sp. gr. 2 No. Boe A = N = ae Origin and remarks. 4° ber. | 7 | 1! 0.941 | —48.6 !_----___- 129.7 | My distillation, Apr., 1907. First 0.45 per cent from 10 kilos of good flowers. | 2] 0.920) —85.5| 1.4846 | 108 B.’s distillate of Aug. 17, 1907. 3 | 0.911 | —89.3 | 1.4840 95 First half of first-quality oil distilled Aug. 17, 1907. 4) 0.921 | —35.1) 1.4821 | 109 | B.’s distillate of Aug. 20, 1907. 5 | 0.939 —34. 2 1.4880 | 131 B.’s distillate of Aug. 22, 1907. 6 | 0.920 —38.9 | 1.4838 98 | B.’s distillate of Aug. 23, 1907. 71 0.925 —45.2 1 -4900 | 110 B.’s distillate of Aug. 24, 1907. Typhoons and poor flowers. ; 8} 0.919 | —45.9| 1.4864} 100 B.’s distillate of Aug. 27, 1907. Rains and typhoons. 9} 0.912 | —48.2| 1.4852 90 B.’s distillate of Aug. 29, 1907. Flowers very poor be- cause of continued typhoons. 10 | 0.922 | —26.0} 1.4794) 117 First quality oil rectified in vacuo. B.’s distillate 90 per cent yield. B.’s distillate of Aug. 31,1907. Flowers so poor because of continued typhoons that 500 kilos were used for 1 | kilo of oil. 12) 0.918 | —88.3/} 1.4808 95 B.’s distillate of Sept. 5, 1907. Flowers good as a result | of three days’ sunshine and hence a large yield of oil. -4898 | 104 B.’s distillate of Sept. 7, 1907. . 4883 | 112 B.’s mixed distillate. | B.’s distillate of Sept. 10, 1907. : 4747 | 102 | Distillate 55 grams obtained by redistilling 100 kilos of ylang-ylang condensation water. 11 | 0.915 | —45.6 | 1.4843 96 ar oO S © as a oo oO wo Bee ra ia 1 or) a S S 17 | 0.917 —39.8 | 1.4785 96 B.’s distillate of Sept. 12, 1907. | 18} 0.922 45.9 | 1.4890] 104 | B.’s distillate of Sept. 3, 1907. | 19 | 0.921 | —40.7 | 1.4825; 108 B.’s mixed distillate. 20 | 0.914! —37.9} 1.4895) 101 | First-quality oil from Mindoro. 21/ 0.949 | —36.1 | 1.4940 | 138 | Distilled in vacuo with steam from selected flowers. 22} 0.827| —42.2| 1.4912 | 126 | B.’s distillate of Feb. 1, 1908, from very good flowers. | 23] 0.958 | —27.0] 1.4910 | 169 0.4 per cent yield from selected flowers with very care- | ful distillation. A very fine oil. YLANG-YLANG OIL. Taste Ill.—Tabulation of the constants of second-grade ylang-ylang oils. Sp. gr. 30° 30° Ester No. 30° ip N ‘p | Bum- Origin and remarks. 4° ber. 1} 0.929 | —69.2 |__--_--__ 84.6 | Corresponds to No. 1, above. Second 0.42 per cent from same distillation as No. 1. 2! 0.910 fo || 89 First half by distillation im vacwo from a provincial oil. By) OSes |) BS) ee 2 Second half by distillation in vacuo of the same pro- vincial oil. Z| OSGTO |) Brad! | 64.1 | Provincial oil from Nueva Caceres. Distillate of Mar., 1907. 5 | 0.912 | —51.5 |--------- 83 My distillation; poor flowers; 0.45 per cent yield. 6} 0.921 | —55.7 |_-------__ 75 My distillation; poor flowers; 0.52 per cent yield. e903 940 ki — So Shee 80 My distillation; poor flowers; 0.61 per cent yield. 8} 0.918 | —42.7 - 4950 84 Provincial oil. 9} 0.918 | —37.2 4954 77.1 Do. 10} 0.942 | —31.3 .4978 87 Corresponds to No. 8, Table II. B.’ssecond-quality oil of Aug. 17, 1907. 11 | 0.918 | —86.0} 1.5000 72 Corresponds to No. 4, Table II. B.’s second-quality oil of Aug. 20, 1907. 12 | 0.917 | —66.7| 1.5032 70 My distillation of second-quality oil from flowers from which the first quality had been previously distilled. Yield, 0.7 per cent. 13} 0.919 | —61.4 | 1.4977 86 Second-quality oil corresponding to No. 5, Table II. 14/ 0.918 | —66.4 | 1.4986 83 Second-quality oil corresponding to No. 6, Table II. 15 | 0.903 | —81.3 | 1.4981 59 | Second-quality oil corresponding to No. 7, Table IT. 16) 0.928 | —30.2) 1.4927 64 | Provincial oil. 17 | 0.918 | —73.5 | 1.4979 80.8 | Second-quality oil corresponding to No. 8, Table II. 18 | 0.906 | —76.0] 1.4991 67 | Second-quality oil corresponding to No. 9, Table II. 19 | 0.926} —75.0| 1.5054 80 | Second-quality oil corresponding to No. 11, Table I. | 20} 0.901 | —44.4 | 1.4935 54 ~ Provincial oil. | 24 | 0.896 | —46.8 | 1.4838 72 | Provincial oil No. 20 rectified in vacuo; 0.65 per cent | yield. 22 | 0.897 | —29.8 | 1.4788 69 Provincial oil No. 20 rectified in vacuo; 0.50 per cent yield. 23) 0.914 | —84.7 | 1.5001 73 Second-quality oil corresponding to No. 12, Table II. 24 0.913 | —66.8 | 1.4926 86 Second-grade ‘‘Sartorius’’ brand. * 25 | 0.910 | —69.0] 1.4972 69 Second-quality oil corresponding to No. 13, Table IT. 26} 0.904 | —87.0| 1.4980 68 Second-quality oil corresponding to No. 15, Table II. 27) 0.922} —85.8 | 1.4962 77 Provincial oil. 28 | 0.926 | —51.6 | 1.5002 89 Do. 29) 0.913 | —84.3] 1.4980 86 Do. 30} 0.915 | —43.3 | 1.4962 81 Iriga oil. Approaches first-grade oil in quality. 31} 0.921 } —36.2 | 1.5002 94 Oil from Guinobatan, Albay. Approaches first-grade oil in quality. 32 | 0.914 | —55.4 | 1.5008 82 Second-quality oil from same locality. Same distil- lation. 33 | 0.920} —42.6| 1.4916 85 | Good second-quality oil from B’s plant. 34 | 0.912 | —45.6 | 1.4928 90 | Oil from Nueya Caceres. Approaches first-grade in | quality. 35 | 0.908 | —80.1 | 1.5082 53 | Oil from Mount Isarog, Ambos Camarines. A very | | poor oil. 386, 0.912 | —32.1} 1.4942 86 Oil from Albay. Very close to first-grade oil. ol | aThe manufacturers of the “Sartorius’’, one of the best brands of ylang-ylang oil, have refused to sell us any of their first-quality oil, and the analysis of number 24 was made from the second quality of the oil, sold in Manila as ‘essence of ylang-ylang We expect to be able to obtain the first-grade oil from European de Pablo Sartorius.” sources, and hope to include this standard brand in our future work. (2 BACON. DISCUSSION OF ANALYTICAL RESULTS. Several regularities will at once be noted. The ester number of first- grade oils is usually 100 or more, whereas that of second grade but rarely rises above 80; the refractive index of the former class is usually 30° D =1.4900, whereas that of the latter ap- proaches 1.5000. his difference is due to the larger content of sesqui- terpenes and resins in second-grade oils, cadinene having a refractive low, being but rarely over N 21)2 2 index N-py =1.5060. The optical rotation of first-grade oils is much lower than that of the second grade, it being but rarely over —45° and usually varying from —32° to —45°, that of the second grade being around —60° and over. This difference is also caused by the high content of the latter in sesquiterpenes. A few provincial oils have a low optical rotation together with a low ester number, and such oils are in general very poor, they are also apt to have a very low specific gravity.” The results all go to show that an oil with a low refractive index, low optical rotations, and high ester number is almost certain to be good, while high refractive index, high optical rotation, and low ester number indicate a second-grade oil. No especial regularities have been noted in the specific gravities of the various oils, save that if an oil has a high specific gravity and high ester content (ester number above 110) it may follow that it may also have a higher refractive index, and oils with all these constants are very superior. (See Table Il, number 23.) Manila buyers of provincial oils are often anxious to ascertain the quantity of flowers used by the distiller im obtaining the oil offered, so that they may judge as to its quality. A number of experiments -were made on the distillation of ylang-ylang oil in vacuo to obtain data on this point and to ascertain whether it might be possible to rectify a lower grade of-oil by such a procedure. One hundred cubic centimeters of first-quality oil were placed in a 200 cubic centimeter, high-necked flask, the distance from the surface of the oil to the exit tube being 16.5 centimeters; a slow, regular distillation was made from a metal bath the temperature of which was kept 15° to 20° hotter than that of the dis- tilling vapor; the total time consumed in the distillation being one hour and twenty-five minutes. te} The original oil~gave the following constants: Specific gravity, a =0:9275 fe} 28n° wads N ee = 1.4883; ester number=117.8. Fraction number 1.—Fifty-two cubic centimeters at 13 millimeters pressure, passing over between 73° and 100°, temperature of metal bath up to 120°. 5 Cf. Table III, numbers 4, 9, 16, 21, 27, 35, all of which numbers were known to represent very poor oils. YLANG-YLANG OIL. 73 A perfectly colorless oil of very good odor but lacking the fine, sweetish, soft flavor of the original. BYaye) 2 ¢ s =(0)921 a= Sills NSE = 14778; ester number=120. Fraction number 2.—Twenty-five cubic centimeters at 13 millimeters pressure, boiling between 100° and 120°, metal bath up to 135°. A water-white oi] having a burnt odor. ° 970 € fo} Specific gravity, 3, =0.916; x =1.4890; At =— 682; ester number=75. Specific gravity, Praction number 3.—V¥ifteen cubic centimeters at 10 millimeters pressure, passing over between 120° and 142°, metal bath up to 155°. ad arm “5 oh0P SOS e aur oOn Specific gravity, —-=0.910; A~-=—97.8; N——-=1.5031; ester number= D D D 109. The residue in the distilling flask was 6 cubie centimeters, N 30=1.5435. A dark brown resin, of rather agreeable odor. Fractions 1, 2, and 3, united, gave 92 per cent of the original oil. This oil was perfectly colorless, but the odor was burnt and not nearly as fine as that of the original oil. i : eee - 30° 30° The constants were: Specific gravity, —— =0.918; A — =—46.5; 30° i —— 4° D 1.4841; ester number=117.1. The greater part of the burnt odor was removed by running air through this oil for some time, but this process did not restore the mild, sweetish odor of the origmal oil. When, during the process of distillation the distillers slightly burn the flowers which they use, the resulting oil is allowed to stand in contact with the air for a day or two, the result being the loss of its burnt odor. I have noted in respect to oils distilled im vacuo that those samples distilled in a hydrogen atmosphere always have more of a burnt, or flat odor than have the ones fractioned with the ordinary air capillary. My experience seems to be that the rectification of oils in vacuo is not an entire success, as the distillates, although ap- parently of the same composition as the oil from which they are distilled, seem to lack in perfuming power; this is especially true of the lasting qualities of the odor. These results suggest that the highest boiling parts of the ylang-ylang oil and even the resins, are very probably im- portant constituents of the whole, possibly they help to fix the more volatile, odoriferous portions. I have always been impressed by the peculiarly lasting fragrance of the resinous residues of the distillation of ylang-ylang oils fractioned in vacuo. The distillation of many ylang-ylang-oils in vacuo has shown that over 50 per cent of the first quality oil will pass over below 100° at 10 millimeters pressure, and when I have tested poorer oils in this respect I have found the amount of substance volatile below 100° at 10 milli- meters which passed over to be proportional to the quantity of flowers used in preparing the oil. Thus one oil distilled from flowers at the yield of 1 kilo for 206 kilos of flowers showed 27 per cent of volatile 74 BACON. constituents under the éonditions named, whereas another prepared in the proportion of 1 kilo to 150 kilos of flowers gave 19 per cent. It follows from this that the distillation test is also of value both in determining the quality of an oil and the proportion of flowers used in preparing it. The only manner in which poor provincial oils may be improved is by redistillation with steam, and this procedure results in large losses. Fractioning with steam in vacuo also seems quite prom- ising, although the process is very slow. Oils thus obtained are quite colorless, and by taking suitable fractions a very fair oil may thus be prepared from a product which before treatment was almost unsalable. The following table illustrates the manner in which the very significant constants of refractive index and ester number vary in the different frac- tions. The numbers represent the successive fractions obtained at L.’s distillery on the dates given: Taste 1V.—Successive fractions of ylang-ylang oil. April 9, 1907. April 11, 1907. September 21, 1907. | No. we Ester we Ester | Specific | 30° Ester > ID) number. D number. | gravity. D number. = 1 1. 4878 163 1.4777 102 0. 927 1. 4888 165 | 2 1.4908 149 1. 4825 185 0. 930 1. 4908 167 | 3 1.1970 105 1. 4906 119 0. 929 1.4945 145 4 1.512 88 1. 4978 91 | 0. 931 1. 5003 105 5 1 1.5085 86 6 al 1.5029 73 7 al 1.5030 61 8 1. 5034 57 9 1. 5023 54 10 | | 1. 5000 49 | A double refined oil (twice distilled) from the same firm gave as fol- lows for the first and second fractions: yay) all Ne =1.4921; specific gravity=0.922; ester number=105. ano me No = 1.4978; specific gravity=0.934; ester number=92. THE ADULTERATION OF YLANG-YLANG OIL. I do not believe that adulteration of ylang-ylang oil is very general in the Philippines. The common adulterants are said to be alcohol, tur- pentine, coconut or other fixed oils, and kerosene. When turpentine is used as an adulterant, it is sprinkled over the flowers and then subjected to distillation with the rest and when small quantities are thus used its detection is exceedingly difficult, as pinene has been reported as a nor- mal, lesser constituent of true ylang-ylang oil. The presence of ter- YLANG-YLANG OIL. 75 penes in ylang-ylang oil probably depends upon the fact that unripe flowers in which terpenes are apt to occur are mixed with those used for distilling. I could not find pinene or other terpenes in 100 cubic centi- meter samples of oils made from fairly good flowers, but the distillation of very unripe flowers gives an oil which has an odor differing entirely from that of ylang-ylang; on the other hand it resembles that of a mix- ture of turpentine and bananas, and doubtless it contains quantities of terpenes and of benzyl or amyl (?) acetate. I have examined a sample of turpentine which was said to be prepared for the use of ylang-ylang distillers; it proved to be a very thoroughly refined, dextro-rotary product, flavored with a trace of essence of pep- permint. If turpentine is present in an ylang-ylang oil in any quantity it gives to the latter a sharp, harsh odor, it lowers the specific gravity, optical rotation and refractive index, and it may be detected in the first fraction upon distilling the ylang-ylang oil in vacuo. If, upon fraction- ing a 100 cubic centimeter sample at 10 millimeters pressure, more than 1 cubic centimeter passes over below 65°, turpentine or some other low- boiling adulterant may be at once suspected. The odor of this fraction will often give some clue as to the adulterant which has been used and if it is suspected that this is turpentine, pinene may be tested for in. the usual manner. The presence of pinene is best proved by its conversion into the bisnitroso-chloride, which with benzylamine gives the corre- sponding nitrol benzylamine melting at 123°. Alcohol is detected in ylang-ylang oil by shaking the sample thoroughly three times with a small amount of water, the latter being thoroughly separated from the oil by centrifugating. The iodoform reaction is then used with the aqueous solution, sodium carbonate and iodine dis- solved in potassium iodide being added. I have satisfied myself that pure ylang-ylang oil gives no reaction with these reagents and that 1 per cent of alcohol can be detected in a 20 cubic centimeter sample by this method. Pure ylang-ylang oil will sometimes give a faint reaction. This is no doubt owing to the alcohol which is used in washing the funnels and flasks in the distillery. Coconut or other fatty oils are detected by the well-known method of placing a drop of the oil on bibulous paper, and this course is satisfactory if the adulterant is present in any quantity. The solubility in 90 per cent alcohol has also been proposed as a test, as fatty oils are soluble with difficulty in alcohol of this strength. I haye found that 3 per cent of coconut oil added to an ylang-ylang oil of the first quality could be at once detected by the opalescence produced by treating the mixture with two volumes of 90 per cent alcohol. However, the test, if used indiscrimi- nately is liable to lead to unreliable results, because a pure, second-grade ylang-ylang oil gives a marked opalescence with alcohol of the same strength; this is due to the fact that sesquiterpenes preponderate in this 76 BACON. quality and the latter are insoluble in 90 per cent alcohol. he dif- ference becomes more marked on using 75 per cent alcohol, as the first quality of ylang-ylang oil dissolves in this strength with only a faint opalescence, while second quality separates in large globules; so that this distinction offers an easy method of roughly judging the quality of the oil. Another method of value is to prepare a 1 per cent solution of the oil in aleohol and compare the odor with a similar one of an oil of known quality, as judgment is much more certain as to the perfuming power when dilute solutions instead of the pure oils are used. One cubic centimeter of each solution can then be poured on separate pieces of bibulous paper, the odor being compared at the end of twelve, twenty- four, or even a longer number of hours; this test gives some idea in regard to the permanence of the odor. Pure ylang-ylang oil obtained by distillation im vacuo leaves a residue of about 5 per cent and of course if fixed oils are present, this will be larger. Moreover, the residue from pure ylang-ylang oil has a refractive Byaye) index as of about 1.5400, whereas a product containing 5 per cent of fo) = =1.5000. Fatty acids can be added coconut oil had an index of N detected in this residue left on distillation by heating it with fused potas- sium bisulphate, for if fatty oils are present the odor of fatty acids as well as a marked one of akrolein is observed. ‘The odor of the fatty acids gives the best sign of their presence in the residue, because pure ylang- ylang oil upon being treated in this way gives a rather sharp odor, which, however, might be mistaken for that of akrolein. If coconut oil has been added to a first-grade ylang-ylang oil to the amount of 5 per cent it can be detected by the odor alone, if the person making the test is familiar with the oil. Petroleum or mineral oil can be detected in ylang-ylang by destroying everything but these adulterants with concentrated sulphuric acid and then distilling the remainder. The use of any adulteration is more emphatically the height of com- mercial folly for ylang-ylang than it is for any other essential oil, for only the product of the highest quality brings a remunerative price. A 10 per cent increase in quantity by means of adulteration may cut the price in two, or may result in an oil which can not be sold at any price. The greatest advance in the ylang-ylang oil industry will take place when the distillers own their groves of trees and can select only ripe, yellow flowers for distillation. This fact is emphasized quite strikingly by the following experiment : Fifty-four and five-tenths kilos (120 pounds) of extra fine flowers, one-half of which were perfectly yellow and ripe, were distilled with steam in the usual manner and the following fractions were obtained: YLANG-YLANG OIL. a =I 30° 30° Number 1: 55 cubic centimeters; specifie gravity, = 0.960 ; A DE es ° No =1.4865; ester number, 178. Dyave) Number 2: 33 cubie centimeters; specific gravity, —~=0.959; AD =—26 5°; RIP . aA ND 1.4914; ester mumber, 160. a: : A on ol? %, 380° Number 3: 90 cubic centimeters; specific gravity, 49 =0.954; A =—34.6°; Net 4956; ester number, 154. Number 4: 80 cubic centimeters; specific gravity, am =0.942; A = —53.4°; 50 2am é Nope 1.5020; ester number. 113. Tubes numbers 1, 2 and 3 united gave the following constants: Specific 2 BY gravity, =O. 958 ; ere rl) s No = 1.4910; ester number, 169. The total oil obtained was 258 cubic centimeters, which is 264 grams, cor- responding to a yield of 0.45 per cent. This yield was nearly twice the normal amount and the quality of the oil was very high, as was shown not only by the analytical figures given above, but also was confirmed by the opinions of Manila experts to whom it was submitted. I believe these experiments indicate that 200 kilos of ripe, yellow flowers will give 1 kilo of a better quality of oil than will 400 kilos of the class of poor, mixed flowers used at the present time. It is a well-known fact of plant physiology that the odoriferous substance is present in the flower in greatest abundance and in finest quality at the time when it is mature and ready for pollination. No doubt, in the course of time much can be done toward improving the yield and quality of ylang-ylang oil by intelligent plant selection. Such work requires much patience and at present there are absolutely no data available save a general opinion that the ylang-ylang trees of the wild mountain regions are not as fragrant as the cultivated ones of the lowlands. Fifty-four and five-tenths kilos of the same flowers were also distilled in a yacuum of 100 millimeters, an exceedingly slow operation. There were obtained 32 cubic centimeters of oil of a very good quality, with the following properties: 97° ic fo} * =36.1; N= 1.4940; ester number, 138. ee aie GAUy Specific gravity, Te es A The low yield is due to the fact that the only apparatus available had the vacuum pipe opening directly into the oil receiver, with only a com- paratively short condenser above, and as a consequence most of the oil was lost by volatilization. 75 BACON. THE COMPOSITION OF YLANG-YLANG OIL. The chemical composition of ylang-ylang oil has been pretty well established through the labors of many chemists. Gal° found no aldehydes or ketones, but benzoic acid as esters of unknown alcohols. Fliickiger,’ correctly names the tree from which ylang-ylang oil is obtained and gives a good historical résumé concerning the oil and its introduction into Europe. He found in the oil a very small amount of benzoic acid, acetic acid and unidentified phenols (from the color reactions with ferric chloride) and suspected the presence of an aldehyde or ketone. He obtained a very small amount of a precipitate with sodium bisulphite. He was unable to identify any of the aleohols from the oil. There would seem to be some doubt as to whether Fliickiger studied a genuine sample of ylang-ylang oil, as the esters of benzoic acid are so abundantly present that the acid is separated in quantity with the greatest of ease. Reychler*® found benzoic and acetic acids, linaloél, geraniol and cadinene. Darzens® found methyl alcohol, para-kresol, benzoic and acetic acids and considered that the para-kresol was present as the acetate, for he states that para-kresol acetate has an odor somewhat like that of ylang-ylang oil. The final and most exhaustive researches on this perfume oil are due to the commercial houses, especially to Schimmel & Company,” whose results are embodied in a patent for artificial ylang-ylang oil. A rational method of analysis of this oil should, if possible, be founded on its composition, and therefore we have undertaken studies in this direction and have succeeded in adding two new substances, formic acid and safrol (isosafrol), to the list of the known constituents of the ylang- ylang oil. Experiment 1—One hundred grams of first-grade ylang-ylang oil were taken; 30° SYN its constants were as follows: Specific gravity, 0.021; Ay = 740-75 30° N D =1.4825; ester number, 108. This oil gave only a very faint reaction with ferric chloride at the contact zone, due to methyl salicylate, showing that the phenols were combined. It is interesting to note that none of the oils I have handled gave a strong color reaction with ferric chloride, and therefore the possibility suggests itself that the ageing of the oil causes a small amount of hydrolysis of the phenol-ethers, for most European observers have obtained pronounced color reactions with ferric chloride. N The oil I used was more than neutralized with 0.1 cubie centimeter of 6 caustie soda. All good ylang-ylang oils are neutral. The oil was heated to 100° in a sealed tube with 15 grams of sodium hydroxide, dissolved in 30 cubic centimeters of water, for eight hours. At the end of this time a considerable amount of solid has separated; there was no pressure in the tube. The contents was dissolved in water and ether, the etherial layer separated, ° Compt. rend. Acad. d. sc., Par. (1873), 76, 1482. "Arch. d. Pharm. (1885), 18, 24. § Bull. Soc. Chim. Paris (1894), 11, 407, 546, 1057. *Tbid. (1897), 27, 83. »eD. R. P. 142, 859 Class 23, a. Schimmel & Company, semiannual report, October 1901 (English edition), p. 53. YLANG-YLANG OIL. 79 the aqueous portion well shaken with ether and the ether several times shaken out with small amounts of water so as to separate the alcohols soluble in that medium, the water being always added to the original alkaline solution. The latter was now distilled with steam and the distillate treated with benzoyl chloride according to Baumann-Schotten, 1.2 grams of methyl benzoate being obtained in this manner. A slight excess of dilute sulphuric acid was now added to the alkaline solution in the distilling flask and the whole was then extracted with ether, the latter being shaken out with small portions of water to remove acids soluble in that medium, these extracts being added to the aqueous por- tions. This was now distilled with steam, the distillate on titration with standard alkali showed 5.54 grams of volatile acids calculated as acetic. The barium and silver salts of the acids were prepared from this distillate and analysed. I. 0.6916 grams barium salt gave 0.6466 gram barium sulphate. Il. 0.5195 grams barium salt gave 0.4872 gram barium sulphate. Calculated (per cent) for Found (per cent). Ba(Cz2H300)22H20 Ba(CHOs)»o I. Ba. 55.06 48.7 60.7 IT. Ba. OO Or, sPemen aia Gite teh saree I. 0.2065 grams silver salt gave 0.1897 gram AgCl II. 0.2050 grams silver salt gave 0.1885 gram AeCl round Calculated (per cent) for (per cent) AgCoH:,02 AgCHO. I. Ag 69.1 64.6 70.58 Il. Ag CE acer Ge ease This aqueous solution of the acids reduces potassium permanganate at once, and readily converts mercuric to mercurous chloride. These results, taken in conjunction with the analytical data and the fact that it was impossible to obtain a white silver salt, leave no doubt but that formic acid was present. The figures obtained for the silver salt are naturally different from those for he barium compound because of the slight reduction to metallic silver caused by the formic acid. The etherial solution containing the acids not soluble in water was now repeatedly extracted with small portions of a solution of sodium carbonate. In this manner there were isolated 9.6 grams of solid acids haying the appearance of benzoic and giving a test for salicylic acid with ferric chloride. The benzoic and salicylic acids were separated by conversion of the salicylic acid into the very insoluble dibromsalicylic acid, 0.6 gram, melting at 221° (found, bromine, 57 per cent; calculated, 57.4 per cent) sufficient water being used to keep the benzoic acid in solution." The remaining acid, after extraction and one crys- 7.7 grams of pure benzoic acid melting at 121° to 122°. After removing the acids, 0.9 gram of phenols having an odor resembling that of isoeugenol and giving a green color with ferric chloride was isolated. A methoxyl determination according to Zeisel gave the following numbers: 0.92 gram of oil gave 0.1405 gram silver iodide equivalent to 0.031 grams of CH;. This would represent 0.99 per cent of the total oil or 0.66 grams CH,OH which would give 2.02 per cent of the total oil as CH,OH. It would require 2.5 grams tallization from ligroin was “Sharpe: Ztsch. f. An. Chem. (1893), 32, 107. 50 BACON. of CH,OH to form the methyl esters with the 9.6 grams of solid acids which were found. The neutral portion of the saponification product distilled in vacuo gave the following fractions: > : 30° , 80° Number 1: 37 grams; B. P. 90° to 120° at 37 mm. A4,=—16.5; N— D =1.478. By aye) Number 2: 15.5 grams; B. P. 120° to 130° at 34 mm.; a” S— sles —— = 1.4797. N D i 302mm Number 3: 17.5 grams; B. P. 130° to 150° at 30 mm.; Kone ee noe’ = 1.4943, D Number 4: 9.5 grams of resinous residue. Treatment of fraction 1 with dilute, aqueous potassium permanganate gives benzoic acid melting at 122°, thus proving the presence of benzyl alcohol in this fraction. i The above results gave the composition of ylang-ylang oil as follows: Component. Per cent. Neutral 81.50 Formie and acetic acids 5.54 Benzoic acid 9.00 Salicylic acid 0.60 Methyl alcohol 2.02 Total 99.56 Several attempts were made to isolate an aldehyde from ylang-ylang oil, as treatment of the oil with fuchsine and sulphur dioxide gave the color change characteristic of aldehydes, but phenylhydrazine shows no trace of reaction with the oil and on shaking thoroughly with freshly prepared and very active sodium bisulphite no such bodies could be isolated by the usual means; so that no more than a trace of aldehydes can be present in ylang-ylang oil. The work with 100 cubic centimeters was repeated with 1,000 grams of ylang-ylang oil in the hope that the decomposition of larger quantities would lead to the discovery of con- stituents heretofore not recognized. Experiment 2.—One thousand grams of good quality ylang-ylang oil from ° Nueva Caceres and haying the following constants: Specific gravity, P= 0.9125 309m. eno cea f AS ae eN De ester number, 100, were used. The oil was saponified in four lots, the first three by using 200 cubic centi- meters of oil and 35 grams of potassium hydroxide dissolved in 150 cubic centi- meters of 92 per cent alcohol the final one of 497 cubic centimeters of oil was hydrolyzed with 60 grams of potassium hydroxide dissolved in 400 cubic centi- meters of the same solvent. Within a few minutes after the alcoholic potash was added the oil became filled with erystals of the potassium salts of the acids. No other solids separate. This was proved by a special experiment with a ioe) YLANG-YLANG OIL. 1 sample of first quality oil in which the solid formed by saponification was filtered, well washed with absolute ether and decomposed by dilute acids, the resulting substances being benzoic, salicylic, formic and acetic acids as well as a very small quantity of phenols. The contents of the first three flasks were completely saponified by two hours’ heating on the steam bath with a reflux condenser, the last fraction by one week’s standing at room temperature. The procedure used in working up the product was like that given in the details of experiment 1. The solid, acid portion: The total amount was 91 grams. The benzoic and salicylic acids were separated by means of their esters., salicylic methyl-ester being soluble in a 10 per cent sodium hydroxide solution. The methyl esters after this separation, boiled almost constantly at 195° and 224°, respectively, so that there would seem to be no reason to suspect any acids other than benzoic and salicylic to be present in the solid, acid portion. The total quantities were 6 grams of salicylic acid and 85 of benzoic. i The soluble acid portion: The alkaline, aqueous solution was first evaporated to a small bulk in order to remove all neutral volatile substances. It was then rendered acid with dilute sulphuric acid, and the volatile acids were distilled. The total volatile acids, calculated as acetic acid, were 63.5 grams. Barium salts were prepared and gave the following analytical data: 0.6995 gram barium salt gave 0.6521 gram barium sulphate Calculated (per cent) for Found (per cent). Ba(C,H30.)2H,0 Ba(HCOs)s Ba 54.97 48.70 60.3 Naturally, the percentage of barium found gives no clue to the relative propor- tions of the two acids present because of the differing solubilities of the salts, the acetate of barium being more soluble in water than the formate. The solu- tion of the soluble acids reduces potassium permanganate and also mercuric chloride, and as neutral salts have been prepared from the solution with a greater percentage content of barium than is in the acetate, the only conclusion is that formic acid is present. The formic acid was estimated according to the method of H. C. Jones” by heating with an excess of standard potassium permanganate in the presence of an alkaline carbonate, then adding a known excess of oxalic acid, and titrating back with potassium permanganate. The result showed that from the original kilo of oil there were separated 41.2 grams of acetic and 17.1 grams of formic acid. A portion of the barium salts of these acids was treated with ethyl alcohol and concentrated sulphuric acid; the resulting esters possessed the characteristic odor of ethyl acetate and formate and no odor was noted suggesting the presence of acids other than those named. Reychler™ states that he has observed a pronounced, rancid odor in the aqueous mother liquors, suggesting small amounts of the higher fatty acids. I could confirm his observation and I believe there is a trace of valerianic acid in the mother liquor. . The phenol fraction—This was in all 10 grams. The phenols gave a green eolor with ferrie chloride and an odor resembling that of kreosol (the 3-methyl ether of homopyrocatechin (C,H,;(CH,)'(OCH,)*(OH)*) was noted in this fraction. There was separated from this fraction according to the method of Baumann- Schotten by the action of benzoyl chloride a small amount of para-kresol ben- Am. Ghem. J. (1895), 17, 539: 18 Toc. cit. 69330 3 82 BACON. zoate, melting at 77° and isoeugenol is also present. I have not attempted thoroughly to study the phenol fraction because of its small amount and because it has been well identified by Schimmel & Company. The neutral fractions after shaking out the alcohol as thoroughly as possible, was dried over anhydrous sodium sulphate, as calcium chloride was difficult to remove, both because of the presence of ethyl alcohol and because of the solid compounds which may be formed with benzyl alcohol and with geraniol. The total neutral oil weighed 808 grams; it was subjected to three careful fractionations in vacuo, using a column of glass beads in a high-necked distilling flask. The fractions obtained, together with their physical constants, are given in the following table:” TABLE V.—I’ractions obtained by distilling the neutral oils left after saponifying 1,000 grams of ylang-ylang oil and removing other constituents. : Boiling point. Sp. gr oe RS enn | Rercenten| Be | D | ND |ierems). |, iow | 15mm, 760 mm. F - CH,+ROMeI. The method is exceedingly convenient after the reagent is prepared, it is quick and uses only about 0.2 gram of oil. It bids fair to displace the older method of obtaining the acetyl number. It is well known that with many aleohols, as for example with linaloédl, the standard methods give an acetyl number many percent too low. This new method promises to give more accurate numbers and is so exceedingly convenient that we are now testing it thoroughly. The results I have obtained in my first series are as follows: Found hydroxyl Calcuted hydroxyl (per cent). (per cent). 10.95 11.4 11.2 11.3 11.1 7, Wilell 11.3 11.4 11.04 11.2 15.8 15.7 15.9 15.6 Linaloiél 11.04 Geraniol ———————, ———__—_——_ Benzyl alcohol The formula used in calculating the percentage of hydroxyl is as follows: X (per cent OH) =0.0764~ where V=volume of CH, at 0° and 760 millimeters and § is the weight of substance taken. Fraction 1 always presented an odor very much resembling that of amyl or hexyl alcohol. By treating this fraction with benzoyl chloride according to the method of Baumann-Schotten and redistilling in vacuo, this odor is removed, the fraction then assumes the odor of terpenes and I was able to obtain a few crystals of pinene nitrosylechloride melting at 103° from this low-boiling portion. This, together with the optical rotation of this fraction would indicate the presence of d-pinene in small quantity. The boiling point of the first three fractions indicates that very little pinene is present. A very small percentage of hydroxyl was also found, so that the major part probably consists of other terpenes. The presence of limonene could not be proved. As these terpenes are of no great importance in determining the odor of ylang-ylang oil, no further attempts were made to identify them. Ww Ber, d. chem. Ges. (1901), 40, 2023. 84 BACON. Fractions 1 to 5 seem to consist of a mixture of terpenes with linaloél and benzyl alcohol, with possibly a small amount of amyl or hexyl alcohol. A special experiment made to separate the terpenes from the alcohols by means of the Grignard method according to the equation CH,MgI+C,,H,,0H —3C,,H,,OMgI+CH, demonstrated that linaloél as well as the alcohols of ylang-ylang oil give addition products which are soluble in ether, so that no separation was possible. Fractions 6 and 7.—These fractions possess an odor much like that of linalodl and in addition another, sweet one, is present. Oxidation of 9 grams of the united fraction with dilute potassium permanganate gave 1.2 grams of benzoic acid melting at 121°. This corresponds to 13.3 per cent of benzyl alcohol. A small portion of these fractions was oxidized with potassium bichromate and dilute sulphuric acid, and by this means the odors of benzaldehyde and of citral were obtained. The fractions 6 and 7 were also heated with finely powdered calcium chloride for one hour on a steam bath and then kept in a cool place for twenty-four hours. A solid compound with calcium chloride separated, this was filtered with the aid of a pump, and well washed with ether. Twenty-nine grams of an oil boiling between 200° and 206° at ordinary pressure were obtained on decomposing this addition product with water. This body had the odor of benzyl alcohol and proved to be the latter almost in its entirety by conversion into benzyl acetate, and also by obtaining an 85 per cent yield of benzoic acid by oxidation. The liquid not entering into combination with calcium chloride, proved itself to be almost pure 2n° linaloél, as is shown by the following constants: Specific gravity, oP = 0.8586 30° 30° eng : é A : D —— Ole: Ny eee boiling point 190° to 195° at ordinary pressure. The odor of this portion was identical with that of linaloél, and citral was produced from it by oxidation with potassium bichromate. Fraction 8—This portion was separated into 10 grams of benzyl alcohol, 11 grams of linaloél, and 3 grams of an oil which solidified with calcium chloride, boiled at a higher point than benzyl alcohol and therefore from its odor and the formation of an addition product with calcium chloride it was determined to be geraniol. Fraction 9.—This fraction, on treatment with powdered calcium chloride im- mediately became hot and solid. This method of separation gave 8 grams of an oil not acted upon by calcium chloride and having a refractive index of fe} No =10181. This portion had the odor of safrol. Oxidization with acid potas- A sium bichromate gave the odor of heliotrope. The oil which was separated from the calcium chloride compound possessed a pronounced geraniol odor, boiled at ° 80° =0:881)'; N——=1.4821. 2D 108° to 115° at 10 millimeters; specific gravity BO. 4° Iraction 10 was similarly separated into 12 grams of geraniol, a small amount ‘of safrol, and into sesquiterpenes. Fractions 11 and 12 consisted almost entirely of sesquiterpenes, among which cadinene is present, as a small yield of cadinene hydrochloride melting at 117° was obtained on treating LO grams of this fraction, dissolved in ether, with dry hydrogen chloride. YLANG-YLANG OIL. 85 SUMMARY OF RESULTS. Ylang-ylang oil is thus seen to contain the following substances: Formic, acetic, valerianic (?), benzoic and salicylic acids, all as esters; methyl and benzyl alcohols; pinene and other terpenes, linalool, geraniol, safrol, cadinene and other sesquiterpenes; eugenol, isoeugenol, p-kresol, probably as methyl ethers; and kreosol. Many of the esters which could possibly be formed by a combination of the isolated alcohols and acids have been prepared in this laboratory in order to become familiar with their properties and especially with their odor. It seems-of interest to give notes in regard to their properties. Benzyl salicylate—No record appears in the literature of the preparation of this compound. It is made as follows: 27.4 grams of salicylic acid are neutralized with 11.2 grams of potassium hydroxide and the solution evaporated to dryness. This potassium salt is heated to 200° for four hours in sealed tube with 25 grams of salicylic acid and 25.2 grams of benzyl chloride. On opening the tube no pressure is observed and all of the benzyl chloride has disappeared. There is obtained 26 grams of a colorless, viscous oil boiling between 186° and 188° at 10 millimeters. The odor is slightly aromatic and pleasant, but not powerful. The oil, after three months in the ice box, has not solidified. Five grams of the ester saponified with alcoholic caustic potash, gave 2.4 grams of salicylic acid melting at 155° (recrystallized from water), and the acid was further identified by conversion into the methyl ester. There also resulted 2 grams of benzyl alcohol, boiling at 204° and converted into benzyl acetate. Benzyl benzoate has been prepared by the action of sodium benzylate on benzaldehyde. The method used in this laboratory was as follows: 50 grams of sodium benzoate, 50 grams benzoic acid, and 43 grams of benzyl chloride were heated on a reflux condenser in a metal bath at a temperature of 200° for two hours. There resulted 46 grams of oil boiling between 315° and 320°. Benzyl valerianate and benzyl butyrate were also prepared by the same general method, the yield being very good. The method formerly in use, which employed the lead salts, did not give good results. Benzyl valerianate can also be prepared in almost quantitative yield by the action of valeryl chloride on sodium benzylate. The properties of this body correspond to those given in the literature. The odor of this ester is not as sweet as is that of benzyl acetate, it being more like that of fruit. It is almost certain that there are traces of benzyl valerianate in ylang-ylang oil. Benzyl-methyl ether is obtained in almost quantitative yield by heating benzyl chloride with a slight excess of sodium methylate dissolved in methyl alcohol, in a sealed tube to 120° for two hours. The transformation is not complete if these same substances are heated only in an oil bath on a reflux condenser, and in addition the benzyl-methyl ether always contains chlorine. . Benzyl-methyl ether is a colorless oil boiling between 166° and 168°; it has an almost nauseatingly sweet odor. There is probably a very small amount of it present in ylang-ylang oil. Benzyl formate—Benzy1 chloride is heated with a slight excess of potassium formate dissolved in formic acid in a sealed tube at 140° for two hours. The body boils at 84° to 85° at 10 millimeters pressure and has an odor sweeter than that of benzyl acetate, but much like it. The yield is over 90 per cent by this method. It is necessary to use a sealed tube, as benzyl chloride does not dissolve in absolute formic acid at 100°. 86 BACON. Geraniol-methyl ether—This body was prepared from the sodium derivative of geraniol and methyl iodide. It is a colorless oil, boiling between 100° and 105° at 10 millimeters and at 208° to 212° at ordinary pressure. Its odor resembles that of geraniol, but it is more like that of grass. Linaloél-methyl ether is prepared in the same manner as geraniol-methyl ether, it boils between 189° and 192°, and its odor is not very different from that of linaloél, it is not as fragrant as that substance. Geraniol benzoate was prepared by the Baumann-Schotten method, using 10 grams of geraniol and 10 grams of benzoyl chloride. The oil boils between 198° and 200° at 15 millimeters. It has quite a pleasant odor, much like that of some of the higher boiling fractions of ylang-ylang oil. SYNTHESIS OF YLANG-YLANG OIL. An attempt was made to prepare an artificial product to test the ac- curacy of these studies on the composition of ylang-ylang oil. The fol- lowing substances were used : Methyl benzoate; benzyl acetate and formate; benzyl methyl ether (trace) ; benzyl valerianate (trace) ; methyl salicylate; benzyl benzoate; cadinene; safrol; isoeugenol-methyl ether; eugenol; kreosol; methyl anthranilate (trace) ; p-kresol- methyl ether; p-kresol acetate. With these ingredients a mixture was compounded smelling deceptively like good ylang-ylang oil. The fluorescence of ylang-ylang oil, which is always present to a greater or lesser degree, is probably due to the pres- ence of methyl anthranilate. The above work, and that of others demonstrates that ylang-ylang oil has a composite odor, derived from that of many constituents. While it is possible to make a very good artificial ylang-ylang oil, I do not believe that distillers of the best quality of ylang-ylang oil have much to fear from this competition, as the odor of a first-class oil seems to litve more permanence than that of the artificial product. This is a result, I believe, of the presence of sesquiterpene alcohols and fragrant resins in the former. i Work on the physical constants and methods of analysis of ylang-ylang oil will be continued as fast as material is available. The future de- terminations, in addition to the constants given in this paper will include the acetyl number, and if possible a phenol number. The acetyl number is undoubtedly of much importance, as is evidenced from the large percentage of fragrant alcohols found in ylang-ylang oil. Our first quality of oil gave an acetyl number of 74, while a second quality only gave one of 42. We will also in our future work use a constant equal to the sum of the ester and acetyl numbers, thus representing the total amount of alcohols and esters in the oils. Such a number for the last mentioned first quality oil was 174 as compared to 110 for the second grade product. We believe we will thus be able from a few simple ana- lytical determinations to draw many conclusions as to the composition of any ylang-ylang oil presented to us and thus be able to judge of its quality. THE COMPOSITION OF HORLICK’S MALTED MILK. By Grorcr F. Ricumonp and W. HE. MusGRAve. (From the Chemical and Biological Laboratories, Bureau of Science.) Since the publication of our monograph * on “Infant Feeding and its Influence upon Infant Mortality in the Philippine Islands” our attention has been directed to errors in the quoted analysis of Horlick’s malted milk given on pages 375 and 378 and also our observation on page 374 that the greatest objection to malted milks as infant foods les in the large amount of insoluble carbohydrates which they contain. The anal- ysis of Horlick’s malted milk therein quoted was taken from Chapin ~ who in turn quotes McGill,’ who reported upon 103 samples representing 22 different brands of infant and invalid foods as found on the Canadian market. His results upon the malted milk in question are the mean of nine separate analyses made by himself and others and are as follows: Per cent. Moisture 2.55 Fat by petroleum ether 1.41 Loss to alcohol and water 63.87 Total albuminoids N x 6.25 14.00 Starch fiber, ash, etc., by difference . 15.68 Ash 3.57 The manufacturers of Horlick’s malted milk maintain, first, that the percentage of fat, 1.41, as quoted is grossly in error and refer to analyses by Smith * who found over six times as much, namely, 8.75 per cent, Chittenden ® who encountered at least 8 per cent, and to European work- ers, including the British Analytical Control and the Glasgow Corpo- ration, who reported 8.85 and 8.8 per cent respectively ; second, that the proteid content, 14 per cent, as found by McGill is also below the true value of 16 to 17 per cent claimed and based upon the analyses covered by the above references; third, that 0.05 per cent represents more nearly the true content of insoluble carbohydrates in their product. Manifestly 1This Jour. Sec. B (1907), 2, 4. * Theory and Practice of Infant Feeding, 2d ed. 1904. * Bull. 59, Laby. Int. Rev. Dept., Canada. ‘Holt on Infancy and Childhood, 1902. ° Dietetic and Hygiene Gazette, 1896. 87 88 RICHMOND AND MUSGRAVE. such variation in the percentages of the important ingredients of an established food stuff are well worth inquiry. In the interest of fairness to the manufacturers and ourselves, therefore, we determined to make a personal examination of the product as found on the Manila market. It hag been known for some time that Adam’s Soxhlett extraction method for the estimation of fats failed in the presence of considerable quantities of carbohydrates and that some other method of procedure was necessary in order to obtain a maximum yield of fat in sweetened con- densed milks, milk powders and malted milks. Modifications of the Soxhlett method such as drying a highly diluted aqueous solution of the material on large surfaces, interstratifying the dried powder with some inert substance such as sand or asbestos or finely grinding the dried powder with powdered glass, before extraction with organic solvents, will increase the yield somewhat. The percentage of fat in Horlick’s malted milk determined in this laboratory both before and after grinding the dried powder with powdered glass was 1.78 and 2.50 per cent respectively. ‘The well-known Babcock volumetric method is also not suitable for such products because of the charring action of concentrated sulphuric acid on the carbohydrates. Leach’s® modification of the Babcock process, which consists in separating the sugars from the fats and proteids by means of copper sulphate before the addition of sulphuric acid, is not very satisfactory when applied to malted milks because of the difficulty of accurately sampling the material and in reading the volume of fat obtained. Several trials on Horlick’s malted milk with Leach’s modification of the Babcock process gave an average of 8.32 per cent of fat. Cochran,’ with a modified Babcock bottle and by the use of equal volumes of 80 per cent acetic acid and concentrated sulphuric acid, claims that the charring action is much less than when concentrated sulphuric acid is used alone. Instead of centrifugating he separates the fat by means of ether, which is evaporated before the volume of fat is taken. Here again the same difficulties in introducing a very hygroscopic powder into the narrow-neck flask and in reading the correct volume of fat are encountered. However, it is only fair to state that our attempts to obtain concordant results with Cochran’s method were made with the ordinary Babcock bottle which does not provide for sufficient means of escape of the ether vapor. In this laboratory much better results were obtained by precipitating the proteids in malted milk with acetic acid and heat and by subsequent extraction of the dry precipitate with petroleum ether. Our exact procedure was as follows: A 1-gram sample of the air-dry powder was transfered to a small breaker, 25 cubic centimeters of water added and the whole stirred to a completely homogeneous solution. It ® Jour. Am. Chem. Soc. (1900), 22, 589. 7 Jour. Am. Chem. Soc. (1905), 27, 906. COMPOSITION OF HORLICK’S MALTED MILK. 89 was then acidulated with 5 cubic centimeters of 1 per cent acetic acid and heated on a steam bath until the albuminoids separated in coarse floccules, after which it was filtered through a weighed platinum Gooch crucible and washed with water until the washings were free from carbo- hydrates. After drying, the crucible and its contents were transfered to a direct extraction apparatus and exhausted with pretroleum ether, The fat was then determined from the loss of weight of the crucible or from the weight of the petroleum ether extract. A mean of four deter- minations by this method gave 8.18 per cent of fat. That all the fat is carried down and retained in the Gooch was evidenced ny negative tests for fat im the clear filtrate. Trillat and Sauton ® have described a new method for the determi- nation of proteids in milk which is identical with the procedure given above for the determination of fat, with the exception that they add 5 drops of commercial formaldehyde to the diluted milk before acidulating with acetic acid, they also extract the dried precipitate with acetone instead of ether. They claim that no proteids could be detected in the filtrate by any of the ordinary reagents. It was hoped that the process as followed would also serve for the estimation of the total albuminoids in malted milks, but it was found that the filtrate contaimed at least one-half of the total proteids of the original powder, furthermore that the fat-free residue left in the Gooch crucible contained about 24 per cent of nonnitrogenous matter calculated on the original weight of malted milk dissolved. The nonnitrogenous matter thus found was not due to incomplete removal of soluble carbo- hydrates, nor was it mineral in nature, for the ash content of the filtrate accounted for the total amount of ash found in the malted milk by direct estimation; therefore, malted milk undoubtedly contains nonnitrogenous organic matter insoluble in excess of boiling acidulated water to that extent. The detailed results of our analysis of Horlick’s malted milk is given in the following table and represents the mean of four separate analyses: Per cent. Moisture (loss at 100° C.) 4.03 Fat (by petroleum ether) 8.18 Total albuminoids nitrogen X 6.25 16.64 Total soluble carbohydrates (loss to boiling water) 64.47 Insoluble nonproteid organic matter, starch fiber, ete. 2.60 Mineral matter 4.08 Summary: First, the comparatively high moisture content is readily explained by the greater relative humidity of this climate; second, the samples examined in this laboratory contain at least 8 per cent of fat. § Bull. Soc. Chim., 39, 906. 90 RICHMOND AND MUSGRAVE. The fat content of a prepared food in which cow’s milk is an important ingredient will be subject to considerable variation, but amounts ap- proximating 9 per cent as reported by others are probably too high, due in most cases at least to approximate volumetric methods of estimation ; third, the malted milk contains 2.6 per cent of nonnitrogenous organic matter insoluble in hot water which reduces the total soluble carbohydrates from 67.63 per cent (the average of three determinations from different sources) to 65.03 per cent which is in close agreement with our results and which more nearly represents the actual amount of soluble car- bohydrates present; fourth, the malted milk im question contains be- tween 16 and 17 per cent of proteids which figures fully substan- tiate the claims of the manufacturers in this respect; fifth, the product contains about 4 per cent of inorganic salts, which figure is also in close agreement with the average of previous findings. EDITORIAL. THE RELATIONSHIP BETWEEN THE EXTERNAL APPEARANCE -AND THE ASH CONTENT OF PHILIPPINE COAL. A quantity of coal from a mine south of Sydney, Australia, was sent to this Bureau in 1907 for test. The coal was fairly compact and for the most part with a hackly fracture. It was markedly characterized by al- ternating dull and lustrous bands parallel to the bedding planes. The lustrous bands were usually not oyer a few millimeters in thickness, while the dull layers were many times that. The luster of the lustrous bands was quite brilliant, approaching the vitreous appearance of obsidian. It occurred to me that the difference in the brilliance of the bands might be due to the content of earthy matter. The two were carefully isolated and the ash determined as follows: Dull portion Lustrous portion (per cent). (per cent). 12.1 2.6 These numbers show that without doubt a large part of the difference in luster is due to the ash content. I do not think that coals from entirely different sources are to be compared, but it is quite probable that the luster of coal from the same vein or same field may vary inversely as the content of earthy matter. T have already called attention’ to the striking similarity between all the coals thus far discovered in the Philippine Islands, and I have arranged the record of a number of samples of these together with some from Australia in the order of their descreasing ash content, in the following table: Philippine coals arranged in order of decrease in ash content. Ash Source of coal, a (per Luster of the coal. cent). Tay apase Msp amy ees Neier ate ee rant Tansey 31.50 | Very dull. Negrossmean Cadiz. aea ees ee ee eae 18.00 | Dull to sublustrous. Batantisl amd |e seas ee eee we es eee nee 14.70 | Dull. WITS a a ap en ee | 14,23 | Dull to very lustrous, uneven. Austral ayes sie ee a ee Ee Sere yen nates! | 12.55 | Dull to lustrous, uneven. Australias —_ -| 12.01 | Dull. | PSL EN ere 9.99 | Dull for the most part. Streak | very lustrous. Negros, Escalante 9.55 | Sublustrous to lustrous. SUT £80 eee eo eee ee 9.06 | Sublustrous, uneven. 4 Given above. 1The Mineral Resources of the Philippine Islands, Bulletin of the Division of Mines, Bureau of Science, Manila, 1908. 91 92 EDITORIAL. Philippine coals arranged in order of décrease in ash content—Continued. Source of coal. a Luster of the coal, | cent). | Philip pinesls] ai spe ee ee ee 7.86 | Sublustrous. Cebuymear Carmen ssase sate eee 7,48 Do. | fe redoetlijay ona US yevelsy | 7.22 | Sublustrous to lustrous. | Philippinewslands)} 22) eee 6.95 | Sublustrous. CebumeariCarmen=s =a eee 6.50 | Sublustrous to lustrous. | TZ OU PRU Za yee ee SE ee ee eae 6.20 | Lustrous. | 120) DUN oan eee eae en ete ey ee ew hl ik 6.00 | Dull to lustrous. Polillo, Vista de Burdeus. 4,38 Sublustrous, uneven. Cebu Mibim g-1oyb 0/2 ene eae 4.17 | Lustrous. laeB artes] ames Ct tis eee seats ann mn Catan 4.12 | Sublustrous, uneven. Cebu) ss 282 a ee eee eet 3.81 | Lustrous. Zamboanga 3.77 Do. Tayabas, Antimonan 3.70 | Dull to lustrous, uneyen. Sam ar So 22s sos Sa ee ene en re cee een ee ee ee 3.49 | Lustrous. Cebu, near Carmen 3.28 Do. Australia's S- 2222s ee ee ee ee 2.60 | Very lustrous. Batan Island, Military Reservation, coal seam No, 5__| 2.50 | Lustrous. | I}, TROON SST apmUAKeR, \WVardoprabayes oe 1.92 | Sublustrous. | Mindoro Bula aca oases iste = a ee ean cnet 1.61 Do. Cebu, mean Carm ec ress sees oe ee aes 1.61 | Lustrous, “Given above. These lusters of these coals were nearly all determined ? at the same time, they are therefore comparable and show that to a marked extent they depend on the content of earthy matter. Perhaps there are also other conclusions which may be drawn from the ash content of a coal. It is generally known that the value of a coal increases with the size and continues until egg size and lump are reached. It is also generally true that the heating power advances in the same manner, but by no means in the same proportion as the above factors. This increase is due, barring physical conditions, mainly to the diminished quantity of ash which may be seen from the following figures of Mr. Somermeier.* The sample was thoroughly air-dried and separated by sifting into various sizes and analyzed by the official method as noted below: ] Volatile | wae | See P c Tot combus- ixe } | Size in fractions of an inch. Moisture. tible | carbon. Ash, | matter. | |i, spy 2G! Aine Eee So See eee een ee 2.05 35, 54 59. 66 2.75 I apy COnply a SO es a ee ee | 1.90 38. 05 58. 40 1.65 1 Sapp EO ath OS eNO AS See EL Lee eae eee eee ee 1.70 38.55 | 58.35 1.40 wy LO |p RI SE Se a ene 1.45 38.80} 58.55 1.20 DEON Sa SS a eS AS pce Oe oe 1.15 39.05 58. 20 1.60 ? This Journal, Sec. A (1907-), 2, 50. *J. Am. Chem. Soc. (1906), 28, 1008. EDITORIAL. 93 The greater percentage of moisture in the fine coal is probably ac- counted for by the phenomena of adsorption. The finer the coal the larger the surface exposed and consequently the greater the quantity of water abstracted from the air and held upon its surface. Autvin J. Cox. STARCH PRODUCTION IN THE PHILIPPINE ISLANDS. The cassava plant (Manihot utilissima Pohl) is found in all parts of the Philippines. The Tagalog name is camoting cahoy. In the northern islands the tubers are extensively used by the natives as a food durmg times of need, while with the Moros it forms a staple article of diet. As all Philippine varieties of the plant contain considerable quantities of hydrocyanic acid, the tubers are not used as extensively for stock food as they should be, for the natives generally do not understand how to treat the plant so as to remove this poisonous acid. The problem of obtaining a good and cheap stock food is an exceedingly important one in the Philippines. Nothing appears to me so promising as cassava and the cowpea. The two plants should be grown together. The greatest demand which cassava makes on the soil is in nitrogen, which the cowpea supplies. By a suitable combination of the cowpea, rich in nitrogenous substances, the cassava roots, rich in carbohydrates, and coconut oil-cake, rich in fats, it is a simple matter to make up a first-class, well-balanced stock food. All these substances can easily be made available in the Philip- pines, so there is no necessity for the importation into the Islands of stock food from foreign countries. If the cassava is planted for its starch, or for alcohol manufacture, some other quick-growing legume such as mungo or peanuts, could be planted, with it. These crops will not only pay well in themselves and add nitrogen to the soil, but they serve the further purpose of keeping out the weeds until the cassava is able to take care of itself. With this introduction as to the best methods of handling cassava, it may be stated that in this plant the Philippines*has the cheapest source of starch in the world, and there is only one other substance which at the present time seems able to compete with it as a source of alcohol, a product of which there is a large native supply, namely, the molasses residue from the crystallization of cane sugar. The cheapest alcohol manufactured on any large scale to-day comes from this source, being made in Cuba and Brazil and sold for 10 cents a gallon. Until the methods of alcohol manufacture from cellulose substances (sawdust, ete.) are perfected, the Philippines has in great quantity the two cheapest raw products for alcohol manufacture. There are no reliable data on the right yield of cassava in the Philip- pines. In Mississippi and Florida, on good ground, 10 tons of roots are 94 EDITORIAL. obtained per acre. The record of yields for this plant found in the literature runs from 4 to 200 tons per acre. In the Philippines the plant has never been raised on a large scale, although several companies are now planting it quite extensively. Plants abount a year old, selected at random from the district in the neighborhood of Zamboanga, Mindanao, averaged 25 pounds of roots each, which, planting 1 meter each way, would give a yield of 50 tons per acre. If 10 tons per acre can be obtained in the Southern States-of America, with a possible growing period of from eight to nine months, it would seem to be perfectly safe to figure as much for virgin Philippine soils, with a growing period of twelve months. One acre of ground in the United States will produce on an average 40 bushels of corn containing 1,500 pounds of extractable starch. One acre of cassava in the Philippines will produce at least 10 tons of roots containing 5,000 pounds of extractable starch. If the fermentable matter is converted into alcohol, the comparison becomes even more favorable to the cassava, as the roots contain in addition to the starch, about 4 to 6 per cent of fermentable sugar, so that from the crop of 1 acre of this plant, over 400 gallons of 95 per cent alcohol could be manufactured. Alcohol can be made from cassava for about the same price as from corn and its manufacture from this substance costs in Peoria, Illinois, $0.032, gold per gallon. One can easily figure the price at which cassava alcohol can be sold and still leave a profit. The cost of manufacturing starch from cassava is also essentially the same as from the potato, and it has been described in a paper soon to appear in the Philippine Agricultural Review by Dr. E. B. Copeland and myself. I shall not treat of it here. Ifa factory for manufacturing starch from the roots is not available, these may be ground up, dried and sifted, the cassava flour resulting haying a starch content of 60 to 75 per cent. Some cassava meal, so prepared by pounding up the roots in a rice mortar and sun drying, has been kept in an open bottle in this laboratory for two years and shows no signs of decay. Dr. H. W. Wiley, of the Bureau of Chemistry, United States Department of Agriculture, quotes a price of 14 cents, gold, per pound for such crude cassava meal. First-class cassava starch will probably average over 3 cents, gold, per pound, and pearl tapioca prepared from it, about one-half cent higher. Heating, or very thoroughly washing, is necessary in preparing cassava starch products so as thoroughly to eliminate all the hydrocyanic acid. This acid is apparently present, combined with other substances, in the form of a glucoside, as we have many times noted that cassava roots which had stood until there was a slight decay, had a very powerful odor of this acid, while no such odor may be detected in the fresh root; hence I would recommend to agriculturists of these Islands that in feeding to hogs, the roots should be ground up and either washed many times with water, or better boiled with water into a starch paste. EDITORIAL. 95 The analyses of Philippine cassava tabulated below, show it to contain about the same percentage of starch as the same plant found in other parts of the world. The following two plants were from a plantation on Basilan Island, the analyses being made by Mr. Reibling: Item. I. II. Ar eloh pl antes sass Weight of roots — The large increase in weight noted for twelve months as compared to ten months is worthy of attention, as I have been repeatedly told by Filipinos in many different parts of the Islands that if the plants are allowed to grow for from eighteen months to two years, the tubers will then be very large, weighing 80 to 120 pounds per plant. Five and three-tenths kilos of cassava of unknown age, from Rizal Province, gave by grinding on a nutmeg grater, 1,350 grams air-dry starch (25 per cent) and 600 grams (11.3 per cent) of fibrous residue ‘containing 64 per cent of starch. Experiment demonstrated that by grinding this fibrous residue dry mm a mortar, a further 3.5 per cent of the total of starch could be obtained. However, under present-day conditions of cheap land and cheap labor in the Philippines, it is not good business policy to attempt to obtain any high extraction of the starch from the roots, as to extract this last few per cent costs relatively more than to remove the first 20 per cent of starch and the money can be used to better advantage in raising more roots. One thousand three hundred grams of cassava roots (said to be 2 years old) were rasped on the machine described in the article referred to above, giving 400 grams of air-dry starch (30.77 per cent) or 27 per cent dry weight, and 110 grams of fibrous residue (8.4 per cent). The starch in this residue was 51 per cent. Four thousand grams of tubers (age unknown) from Batangas Proy- ince, gave 1,050 grams of dry starch (26.2 per cent), and 345 grams of fibrous residue (8.5 per cent). The aboye percentages are quoted as samples of the yield of starch which may be obtained in a commercial way from Philippine camoting cahoy. Other analyses made in this laboratory run from 24 to 50 per 96 EDITORIAL. cent of starch and it seems reasonable to assume that 25 per cent of air- dry starch (14 to 18 per cent water) may be commercially extracted from the plant. Samples of arrowroot (Maranta arundinacea Linn.) grown in the Islands contained from 18 to 22 per cent of starch. The plant is raised only as a food for hogs. It seems rather extravagant to feed to hogs one of the highest priced of starches. Sincamas (Pachyrhizus bulbosus Britton) |P. angulatus| tubers gave 2.5 to 10 per cent of commercially extractable starch, according to the age of the plant, the lowest yield being obtained from tubers 24 months old, and the highest from those 12 months old. Tacca pinnatifida Forst., yielded 22.3 per cent of starch. This plant is rasped very easily and the starch is more easily obtained in a pure state from it than from any plant I have handled. Tacca starch sells for a higher price than the others, being called in the world’s market Bermuda arrowroot. Dioscorea sp. gave 11 per cent of commercially extractable starch with a total starch content of 14.3 per cent. This starch is remarkable for the small size of its granules. The seeds of Cycas circinalis Linn., which are sometimes used as a source of “sago”, yielded 31.2 per cent of starch. The tubers of Amorphophallus campanulatus Blume are very large, but from them we were only able to obtain as the highest yield 4.5 per cent of starch. The presence of numerous spicules of calcium oxalate renders the preparation of an edible starch from this plant very difficult. Raymonp F. Bacon. ILLUSTRATIONS. PLATE I. Fic. 1. Starch from Manihot utilissima Pohl. 2. Same in polarized light. 3. Starch from Tacca pinnatifida Forst. Prate II. Fie. 4. Starch from Vacca pinnatifida Forst, in polarized light. . Starch from Tacca pinnatifida Forst. 6. Same in polarized light. OL PrateE III. Fig. -I . Starch from Dioscorea sp. In polarized light, this starch shojvs no change. ) . Starch from Cycas circinalis Linn. . Same in polarized light. 69330 ES ee 97 co CO (PHIL. JouRN. Scr. Vou. III, No. 2. BACON: STARCH PRODUCTION. ] PLATE |- [PuHtIL. JourN. Scr. Vou. III, No. 2. STARCH PRODUCTION. | BACON : Fic. PLATE BACON: STARCH PRODUCTION. ] [PuHIuL. Journ. Scr. Vou. III, No. 2. Fic. 9. PLATE Il!I1. THE ASCENT OF MOUNT PULOG. The highest mountain in northern Benguet which has ever been ascended so far as is known, is Mount Pulog, recently clmbed by Mr. Charles G. Benson and party of the Bureau of Lands. The notes which follow are taken from Mr. Benson’s account of the trip. Mount Pulog is in northeastern Benguet not far from the line between Nueva Vizcaya and Benguet. MKabayan, the settlement from which the start for the mountain was made, is a journey of a day and a half or two days from Baguio. From Kabayan the party went by the regular trail to the barrio of Lutab. One-half mile south of Lutab they turned off on the old Spanish horse trail which runs higher up on the hills than the trail at present used, and followed it to the Adat River. From this point they took a foot trail which runs up the canon of the river at an average height of about 90 meters above it. After following this for one-half mile they traveled in an irrigation ditch for approximately two miles, then descended to the level of the river, crossing it at a point where two branches, one coming from the south of Mount Pulog and one from the north of that mountain, unite. After crossing the fork from west to east they climbed straight up over a very difficult foot trail to Ankiki, a little [gorot barrio of about four families, at an altitude of approximately 2,190 meters above sea level. The trail from Ankiki to the top of Mount Pulog runs around the base of the main peak and over the tops of two subsidiary ones, after which it descends to the rancheria of Tinuk or Tinak, which is about 1,520 meters below the top of the mountain and lies to the south of the Asin Grande basin. The top of Mount Pulog for a distance of about 240 meters below the summit, was found to be covered with very coarse-bladed grass a foot high. The height of the mountain, carefully estimated from barometric read- ings, 1s 2,890 meters. Ice five-eighths of an inch thick formed 60 meters below the summit during the night that Mr. Benson and his party spent there. A sufficient quantity of dead pine-wood for camp fires was obtained near the camp, 90 meters below the summit. The time occupied in travel between the several points on this trip was approximately as follows: Kabayan to Lutab, ninety minutes; Lutab to the first barrio, one hour; the first barrio to the river bed, one and one-half hours; the river bed to Ankiki, four hours; Ankiki to the summit of Mount Pulog, two hours. 99 100 EDITORIAL. The people of the region passed through by this party are Benguet Igorots, but the following differences were noted between them and the main body of the people of the same tribe. The houses at Ankiki, while similar to those of others of this tribe, were rather better built, haying sides of boards. The language spoken by the people of this barrio was hardly intelligible to an interpreter who belonged to the Ibaloi division of the Benguet Igorots. The people of Ankiki dress like the other Ben- guet Igorots. They are called by the latter Kadasdn which is said to mean the people who live where the oak trees grow. Their only agri- cultural product is camotes. They keep hogs, dogs and a few chickens. They are great hunters and kill large numbers of deer and wild hogs. The people of the barrio of Tinuk are called Busols by the Benguet people. This name is practically meaningless, as it is the common designation for people who seem to the Benguet Igorots more wild and uncivilized than themselves. The people of Tinuk raise rice but do not terrace the hillsides to any considerable extent. Merron L. Miter. i : EDITED BY “y, PAUL C. FREER, M. D., Pu. D. { ° WITH THE COOPERATION OF f MERTON L. MILLER, Ps. D.; GEORGE F. RICHMOND, M. S. W. D. SMITH, M. A.; A. J. COX, Pu. D. RAYMOND F. BACON, Pu. D.; CHARLES S. BANKS, M. S. H. D. GIBBS. B. S.; R. C. McGREGOR, A. B. THE BUREAU OF SCIENCE GOVERNMENT OF THE PHILIPPINE ISLANDS | A. GENERAL SCIENCE MANILA _ BUREAU OF PRINTING 1908 PREVIOUS PUBLICATIONS OF THE BUREAU OF GOVERNMENT LABORATORIES, 1No. 1, 1902, Biological Laboratory.—Preliminary Report of the Appearance in the Philippine Islands of a Disease Clinically Resembling Glanders. By R. P. Strong, M. D. No. 2, 1902, Chemical Laboratory.—The Preparation of Benzoyl-Acetyl Peroxide and Its Use as an Intestinal Antiseptic in Cholera and Dysentery. Preliminary Notes. By Paul C. Freer, M. D., Ph. D. 1Wo. 3, 1903, Biological Laboratory.—A Preliminary Report on Trypanosomiasis of Horses in the Philippine Islands. By W. E. Musgrave, M. D., and Norman E. Williamson. 1No, 4, 1903, Serum Laboratory.—Preliminary Report on the Study of Rinderpest of Cattle and Carabaos in the Philippine Islands. By James W. Jobling, M. D. 1WNo. 5, 19038, Biological Laboratory.—Trypanosoma and Trypanosomiasis, with Special peter ends to Surra in the Philippine Islands. By W. HE. Musgave, M. D., and Moses . Clegg. 1No. 6, 1908.—New and Noteworthy Plants, I. The American Element in the Philip- pine Flora. By Elmer D. Merrill, Botanist. (Issued January 20, 1904.) 1No. 7, 1908, Chemical Laboratory.—The Gutta Percha and Rubber of the Philippine Islands. By Penoyer L. Sherman, jr., Ph. D. 1 No. 8, 1903——A Dictionary of the Plant Names of the Philippine Islands. By Elmer D. Merrill, Botanist. 1No. 9, 1903, Biological and Serum Laboratories—A Report on Hemorrhagic Septi- cemia in Animals in the Philippine Islands. By Paul G. Woolley, M. D., and J. W. Jobling, M. D. 1No. 10, 1908, Biological Laboratory.——Two Cases of a Peculiar Form of Hand Infection (Due to an Organism Resembling the Koch-Weeks Bacillus). By John R. McDill, M. D., and Wm. B. Wherry, M. D. 1No. 11, 1903, Biological Laboratory.—Entomological Division, Bulletin No. 1: Prelimi- nary Bulletin on Insects of the Cacao. (Prepared Hspecially for the Benefit of Farmers.) By Charles S. Banks, Entomologist. 1WNo. 12, 1903, Biological Laboratory.—Report on Some Pulmonary Lesions Produced by the Bacillus of Hemorrhagic Septicemia of Carabaos. By Paul G. Woolley, M. D. No. 18, 1904, Biological Laboratory.—A Fatal Infection by a Hitherto Undescribed Chromogenic Bacterium: Bacillus Aureus Fetidus. By Maximilian Herzog, M. D. 1No. 14, 1904.—Serum Laboratory: Texas Fever in the Philippine Islands and the Far East. By J. W. Jobling, M. D., and Paul G. Woolley, M..D. Biological Laboratory: Entomological Division, Bulletin No. 2: The Australian Tick (Boophilus Australis Fuller) in the Philippine Islands. By Charles §. Banks, Entomologist. No. 15, 1904, Biological and Serum Laboratories —Report on Bacillus Violaceus Ma- nile: A Pathogenic Micro-Organism. By Paul G. Woolley, M. D. 1 No. 16, 1904, Biological Laboratory.—Protective Inoculation Against Asiatic Cholera: An Experimental Study. By Richard P. Strong, M. D. No. 17, 1904.—New or Noteworthy Philippine Plants, II. By Elmer D. Merrill, Botanist. iWo. 18, 1904, Biological Laboratory.—I. Amebas: Their Cultivation and Etiologic Significance. By W. E. Musgrave, M. D., and Moses T. Clegg. Il. The Treatment of Intestinal Amebiasis (Amebic Dysentery) in the Tropics. By W. BE. Musgrave, M. D. No. 19, 1904, Biological Laboratory.—Some Observations on the Biology of the Cholera Spirillum. By W. B. Wherry, M. D. No. 20, 1904.—Biological Laboratory: I. Does Latent or Dormant Plague Bxist Where the Disease is Endemic? By Maximilian Herzog, M. D., and Charles B. Hare. Serum Laboratory: II. Broncho-Pneumonia of Cattle: Its Association with B. Bovisepticus. By Paul G. Woolley, M. D., and Walter Sorrell, D. V. S. III. Pinto (Pao Blanco). By Paul G. Woolley, M. D. Chemical Laboratory: IV. Notes on Analysis of the Water from the Manila Water Supply. By Charles L. Bliss, M.S. Serum Laboratory: V. Frambesia: Its Occurrence in Natives in the Philippine Islands. By Paul G. Woolley, M. D. No. 21, 1904, Biological Laboratory—Some Questions Relating to the Virulence of Micro-Organisms with Particular Reference to Their Immunizing Powers. By Richard P. Strong, M. D. No. 22, 1904, Bureau of Government Laboratories.—I. A Description of the New Build- ings of the Bureau of Government Laboratories. By Paul C. Freer, M. D., Ph. D. IT. A eC aeleeue of the Library of the Bureau of Government Laboratories. By Mary Polk, ibrarian, 1No. 23, 1904, Biological Laboratory.—Plague: Bacteriology, Morbid Anatomy, and Ha (Including a Consideration of Insects as Plague Carriers). By Maximilian erzog, M. D. No. 24, 1904, Biologicat Laboratory—Glanders: Its Diagnosis and Prevention (Together with a Report on Two Cases of Human Glanders Occurring in Manila and Some Notes on the Bacteriology and Polymorphism of Bacterium Mallei). By William B. Wherry, M. D. No. 25, 1904.2—Birds from the Islands of Romblon, Sibuyan, and Cresta de Gallo. By Richard C. McGregor. No. 26, 1904, Biological Laboratory.—The Clinical and Pathological Significance of Balantidium Coli. By Richard P. Strong, M. D. No. 27, 1904—A Review of the Identification of the Species Described in Blanco’s Flora de Filipinas. By Elmer D. Merrill, Botanist. No. 28, 1904.—1I. The Polypodiacee of the Philippine Islands. II. Edible Philippine Fungi. By Edwin B. Copeland, Ph. D. No. 29, 1904.—1. New or Noteworthy Philippine Plants, III. II. The Source of Manila HElemi. By Elmer D. Merrill, Botanist. No. 30, 1905, Chemical Laboratory.—I. Autocalytic Decomposition of Silver Oxide. II. Hydration in Solution. By Gilbert N. Lewis, Ph. D. No. 31, 1905, Biological Laboratory.—I. Notes on a Case of Hematochyluria (Together with Some Observations on the Morphology of the Embryo Nematode, Filaria Nocturna). By William B. Wherry, M. D., and John R. McDill, M. D., Manila, P. I. II. A Search Into the Nitrate and Nitrite Content of Witte’s ‘‘Peptone,”’ with Special Reference to Its aermuence Ren rane Demonstration of the Indol and Cholera-Red Reactions. By William B. erry, M. 1 Out of print. 2 The first four bulletins in the ornithological series were published by the Ethnological Survey under the title ‘Bulletins of the Philippine Museum.” Ldter ornithological publications of the Government appeared as publications of the Bureau of Government Laboratories. (Concluded on third page of cover.) THE PHILIPPINE JOURNAL OF SCIENCE A. GENERAL SCIENCE Vou. II JUNE, 1908 No. 3 METHYL SALICYLATE 1.—THE SEPARATION OF SALICYLIC ACID FROM METHYL SALICYLATE AND THE HYDROLYSIS OF THE ESTER. By H. D. Gress. (From the Chemical Laboratory, Bureau of Science, Manila, P. I.) Since salicylic acid and the salicylates have been prohibited in foods,* it becomes necessary in many cases to separate salicylic acid and its metal salts from its esters. The methyl ester, either the synthetic preparation or oil of gaultheria, or oil of betula, is often found to be a constituent of many non-alcoholic beverages, such as the so-called root beers, sarsaparillas, and soda-water flavors. The United States Pharmacopceia and the National Formulary * authorize its use as a flavoring agent, and it is therefore often found in emulsions, the most common of which is cod-liver oil and other pharma- copceelal preparations. Salicylic acid or its salts and its methyl ester may be, and often are, found together in the above preparations ; first, through the incorporation of both in the original mixture; second, when methyl salicylate, or oil of gaultheria, alone is used the ester may contain varying amounts of free salicylic acid as an impurity; third, when a comparatively pure ester is employed, free salicylic acid may subsequently become a constituent of the compound through the hydrolysis of the ester. Regarding the first of these sources, it is sufficient to note that pre-— servatives of various kinds, borax and boric acid, benzoic and salicylic 1U. S. Dept. Agric., Food Inspection Decision 76 (1907). 23d ed. (1906), 46. 71978 : 101 102 GIBBS. acids, have been found by the writer and other investigators in soda-water flavors, root beers, sarsaparillas and cod-liver oil emulsions, both when methyl salicylate was present and absent, and several manufacturers have verified the findings by submitting their formulas for some of these prep- arations. Im many cases it is possible that a preservative, in addition to the methyl salicylate, is quite superfluous, the ester probably having antiseptic qualities * sufficient to render the employment of other steriliz- ing agents or processes unnecessary. Concerning the second source of salicylic acid, namely, as an impurity in the methyl salicylate, an examination of all of the different samples available in this laboratory and in the city of Manila, eight in all, has revealed the presence of the free acid in every case. ‘Two of these samples were represented to be genuine oil of gaultheria, and six were synthetic preparations. All were of European exportation and had been in stock in this city from a few days to over a year. ‘The amounts of free salicylic acid varied from a trace in one laboratory sample to 0.025 per cent by weight in a genuine oil of wintergreen. ‘These small amounts do not wholly account for the larger quantities of salicylic acid or its salts which have been found in a number of different preparations upon the local markets and entering the port of Manila. The third source, namely, the hydrolysis of the ester, will be shown * to account, in many cases, for the presence of free salicylic acid in prep- arations in which comparatively pure methyl salicylate has been employed as an ingredient. With alkalies the rate of hydrolysis is very rapid; it is slower with acids, and even with distilled water the hydrolysis is measurable. The temperature is an important factor of the rate. It is therefore not surprising that the formation of salicylic acid from methyl salicylate in this way is quite appreciable in foods or drugs which have been shipped by vessels to this port. The temperature of the holds of the vessels often rises above 30° in the tropics. The voyage by fastest steamers from Hurope or the United States occupies about one month and by sailing vessels a number of months, and during the entire voyage the rolling and pitching of the vessel produces a constant agitation of the contents of bottles, casks and other containers, maintaining, in all, favorable conditions for hydrolysis. THE DETERMINATION OF SALICYLIC ACID IN METHYL SALICYLATE. The free acid can be titrated directly. The indicators which have been found to be applicable are Congo red and erythrosin. Alfred J. Cohn * says, “Congo red may be used for estimating mineral acids in the *It is hoped that this investigation will form a part of a later paper. * While this phase of the question will be touched upon here, it will be further dealt with in a later paper. * Indicators and Test Papers, J. Wiley & Sons, New York (1904), 56. METHYL SALICYLATE IT. 103 presence of organic acids, as the latter do not affect it.” This has been found to be an error, as salicylic acid can be accurately titrated, the end point being very sharp when either standard sodium hydroxide, carbonate or bicarbonate solutions are used, the carboxyl group only being affected. Walker and Wood ° have used Congo red for titrating salicylic acid against baryta. Hrythrosin has also been found to give fairly good results, although Congo red has been used almost entirely throughout this work. In titrating the free acid in methyl salicylate, from 5 to 20. cubic centimeters of the ester are shaken with an equal quantity of neutral, distilled water in a glass-stoppered flask, and standard alkali, a added . until the color indicating the end point remains permanent on shaking. Standard solutions of sodium acid carbonate’ are best used in this titration, for reasons explained further on, although sodium hydroxide solutions give accurate results. The titrations were carried out at the room temperature, which varied in this laboratory from 28° to 34°. In order. to show that the acidity of the samples was not due to acids other than salicylic, the following method was employed: Ten cubic centimeters of the ester or oil of gaultheria were extracted three times with 5 cubic centimeter portions of — sodium acid carbonate. The acid } 10 carbonate solutions were united, extracted three times with chloroform to remoye the ester which was in solution, made acid with sulphuric acid (1 to 3) and extracted three times with chloroform. The chloroform extracts were united, filtered into a weighed dish, and evaporated spon- taneously in a vacuum desiccator. After weighing the residue, it was dissolved in hot water and the salicylic acid determined colorimetrically.® TABLE I.—Amounts of salicylic acid in natural and artificial oil of gaultheria. Salicylic acid— Sample. Amount. ; By titrat- Beate Colorimet- ing. By weighing.® tically. ce. Per cent. Per cent. Oil of gaultheria (genuine) ---_____ 10 0.025 | 5.5mg.—0.046 per cent__ 0.028 Synthetic? 2S aaa enn 10 0.0113 | 3.9mg.=0.033 per cent__ 0. 0113 57. Chem. Soc. (1898), 73, 619. 7 Standard solutions were made from Kahlbam’s sodium acid carbonate, which was found to be very pure. * Methods of Analysis, Bull. U. S. Dept. Agric. (1907), 107, 180. "The weights of the salicylic acid are evidently too great for the reason that drying was imperfect. Small quantities of the acid are so easily volatilized that it was considered preferable to err in the opposite direction and rely upon the colorimetric method for accuracy. 104 GIBBS. SEPARATION AND DETERMINATION OF SALICYLIC ACID AND METHYL SALICYLATE IN FOODS AND DRUGS. The substance under investigation, containing salicylic acid and methyl salicylate, is made strongly alkaline to Congo red with an approximately normal solution of pure sodium acid carbonate, free from normal car- bonate’? and, if not homogeneous, the aqueous solution is separated and the process repeated with the residue until it is thoroughly extracted by the sodium acid carbonate solution. All of the salicylic acid has now passed into the acid carbonate solution in the form of sodium salicylate together with small amounts of methyl salicylate. This solution is extracted repeatedly, not less than three times, with small amounts of chloroform™ until all traces of methyl salicylate have been removed. The sodium acid carbonate solution is now made acid with sulphuric acid (1 to 3) and extracted in the usual way to remove and determine the salicylic acid.*? | | This method has been successfwly applied to emulsions of cod-liver oil which are usually very difficult to separate. The sodium acid carbonate layer, carrying the salicylic acid and small amounts of methyl salicylate, can be separated in a rapidly revolving centrifuge. With non-alcoholic beverages and soda-water flavors, the method is especially easy of manip- ulation. During the process of extraction, while the methyl salicylate is still in the solution with the salicylic acid salts, the temperature should not be unduly raised for the reason that the rate of hydrolysis of methyl salicylate is accelerated with increase in temperature. During the manip- ulation in this laboratory, where the temperature is always high, the solutions have been kept below 35°, which temperature has been found to be a fairly safe limit. Lower working temperatures are, of course, to be desired. The ester, separated by chloroform extraction,” is saponified by heat- * Solutions of sodium acid carbonate lose carbon dioxide and therefore should be freshly prepared and kept in well-stoppered bottles. The loss of carbon dioxide, the increase of normal sodium carbonate, and consequent increase of sodium hydroxide in the solution is in most eases counterbalanced by the acidity of the substance under examination. When this substance is very acid it is best made alkaline by the addition of solid sodium acid carbonate in order to avoid a great increase in the bulk of the solution. “Chloroform has been found to be better than ether for removing the methyl salicylate from this solution, for the reason that it is less miscible with the aqueous solution. = Methods of Analysis, Loc. cit. *Tn the case of oil emulsions and some other mixtures the ester is best separated, after the removal of the salicylic acid, by steam distillation from a sulphuric-acid solution. Since methyl salicylate is partially hydrolyzed on heat- ing in a sulphuric-acid solution, it is necessary to carry on the distillation until all of the salicylic acid formed has passed over into the receiver. METHYL SALIGYLATE I. 105 ‘ ing in a flask with reflux condenser attached, on a steam bath, with a large excess of strong caustic alkali solution. After saponification is complete, half an hour usually being sufficient, the condenser is detached and the heating is continued until all of the chloroform is expelled. The solution is then diluted to a known volume and the salicylic acid determined in aliquot portions. The following quantitative experiments serve to show the manipulation and the accuracy of the method. 1.0256 grams methyl salicylate were dissolved in 50 cubic centimeters of chloroform and 10 cubie centimeter portions saponified with 10 cubic centimeters of a 25 per cent solution of caustic potash. After evaporation of the chloroform, the residue was diluted to 100 cubic centimeters and 2 cubic centimeter portions ~made acid with sulphuric acid (1 to 3) and extracted four times with small amounts of chloroform. The chloroform was evaporated in a vacuum desiccator, and the residue dissolved in 100 cubic centimeters hot water. The salicylic acid determined colorimetrically “ in this solution gave 1.0640 grams methyl salicylate. 1.2277 grams treated as above gave 1.2667 grams. 0.1568 grams dissolved in 10 cubic centimeters of a 25 per cent solution of sodium hydroxide gayve— Us Il. 0.1499 gram. 0.1565 gram. THE HYDROLYSIS OF METHYL SALICYLATE WITH SODIUM CARBONATE AND SODIUM HYDROXIDE.!° Solutions of sodium hydroxide, approximately N and 7 were made >) by dissolving clean, metallic-sodium in distilled water from which the gases had been expelled by boiling. These were agitated in bottles with an excess of methyl salicylate and 10 cubic centimeter portions were removed and titrated at intervals.‘° The reactions were all carried on at 30°, with variations not exceeding +1°. ‘This is the prevailing temperature in this locality. In the following tables, ¢ is the time expressed in hours, v the volume = of = sulphuric acid used to neutralize 10 cubic centimeters of the reac- tion solution at time ¢, and @ is the percentage of sodium hydroxide which has been used in the reaction. “Color comparisons made with a wedge colorimeter. More extended investigation of the hydrolysis of methyl salicylate with acids, alkalies and water and the catalytic action of tropical sunlight is being carried on and will probably be presented in a later paper. The cases of sodium carbonate and hydroxide are here taken up merely to show the basis of the analytical methods. © Ostwald-Luther: Physiko-Chemische Messungen, Leipzig (1902), 447. 106 GIBBS. Tarte Il.—Hydrolysis of methyl salicylate by sodium hydro«ide. 0.203 normal NaOH; T = 30° + 0.5°. t v u t v x 0 20. 30 0. 00 2.30 88. 67 1 14.70 27.58 7 1.70 91. 62 2 9. 60 52.70 8 1.25 93.84 3 6. 65 67.24 24 0.30 98. 51 4 4. 60 77.34 31 0. 20 99. 01 5 3.30 83.74 11 0.30 96.96 |) 18 0.15 98. 48 15 20 97.97 35 -10 99. 00 | IN are : ~~ sodium carbonate solutions were made from the pure salt and also from the carbonate formed by the ignition of sodium oxalate. A large excess of the ester was used in every case. TaBLE II].—AHydrolysis of methyl salicylate with sodium carbonate. FIRST SERIES. 0.2 normal NasCO3; T = 30° + 1°. t v x | t v x 0 | 20.0 0.0 | 32 | 14.20 | 29.0 t | 19.4 BO | 48 | 13.2 34.0 2 | 19.1 Ab5iillWemu ogi 1280p 11385 3 | 18.7 6.5 | Dele 2sS5 eile e38425 5 | 18.1 9.5 | 80 | 12.1 39.5 | 7 |) 1765 | 75 |< 125 | 4.46 | 42.7 8 | 17.3 13.5 144 | 11.28 | 43.6 | 94 | 14.9 95.5. || 152 | 11.24 | 43.8 28 | 14.4 FO cleo mee ah, 2 44.0 31 | 14.25 | 98.75 || 194) 11.1 44.5 } SECOND SERIES. g.| 17.5 | 125 || 64 | 12065 || 86.75 i Gye) aged 5 | 80 aPANGY | Bi. PE} 10 | 17.0 0 sg | 11.97 | 40.15 | 11 | 16.8 O | 138 | 11.40 | 43.0 13°] 16:5 5 |, 12 | 11.2 44.0 14 | 16.2 Ob |) GO| ieee 2) ea 3n0) 16 | 15.9 5 | 176 | 11.18 | 44.1 $ 32 | 14.25 .75 || 202 | 1102 | 44.9 33° | 14.1 99.5 || 295 | 11.0 45.0 36 3.9 | 30.5 | 254 | 10.95 | 45,95 39 | 13.75 | 31.25 323 | 10.7 46.5 40 | 15.7 | 381.50 4s9 | 10.45 | 47.75 56 | 12.9 35.5 METHYL SALICYLATE I. 107 Taste Il].—Hydrolysis of methyl salicylate with sodium carbonate—Continued. THIRD SERIES. 16 15.8 21.0 | 96 11.9 40.5 w | ise | a7 || an | 13 43.5 18 | 15.5 | 1G) || THY 44.0 21 | 15.25 | 23.75 || 168 | 11.2 44.0 93 | 15.1 24.5 |) 184 || Wa || 44.95 24 | 14.9 25.5 || 210 | 11.0 45.0 40 } 13.7 31.5 || 233 | 10.9 45.5 | 44 | 13.4 | 33.0 | 262 | 10.88 | 45.6 | 48 | 13.3 33.5 || 497 | 10.40 | 48.0 | Ge bh 36.5 || 624 | 10:2 49.0 | 72 | 12.45 | 37-75 || 665 | 10.18 | 49.35 ss | 12.0 40.0 || 737 | 10.00 | 50.00 loo 80 inal Q yy N a x { BS | KR = Ee s I : | x x + 2 | . 40 S 5 1R G 8 as 4 ai ea oe & iy i TL a 20 | te | tt A N HOUR|S al LOG. HOURS o 20 4O 60 80 100 2 3 737 HOURS Fie. I—Hyprotysis or Merayn SALICYLATE. The rate of hydrolysis with sodium carbonate is a smooth curve, the break in the diagram being due to the change of scale. The curves shown in fig. 1 are constructed from the above tables. It is to be noted that the hydrolysis of the ester with sodium hydroxide goes to completion; that is, to the point where all of the hydroxide has been used in the reaction, or at least it goes very nearly to completion in about twenty-four hours. With sodium carbonate, equilibrium, for all practical purposes, is reached in about one month, at the point where all of the 108 GIBBS. normal carbonate has been converted into the acid carbonate according to the equation : 7 C,H, . OH . COOCH,+Na,C0,—C,H, . OH . COONa-_CH,OH+ NaHCO, To prove that this is the end point of the reaction, or at least the point where the rate is exceedingly slow, the ester was shaken for days with pure N y : arts , i0 sodium acid carbonate t* solution in a number of sealed tubes. While a slight reaction was noted, it is believed that the substances were practically in equilibrium.t® Any reaction taking place is not suf- ficiently rapid to affect the accuracy of the analytic methods previously described, which depend upon sodium acid carbonate for the removal of salicylic acid as sodium salicylate from the presence of the methyl ester, without saponification of the latter. Cahours *° says that concentrated solutions of alkalies react with methyl salicylate in the cold to produce the salts of the ester. Freer ** has prepared sodium salicylic ethyl ester by the action of sodium upon the ester and by the action of sodium hydroxide upon the ester in etherial solution. He mentions the fact that the compound thus formed is easily hydrolized by moisture. The reactions with dilute solutions of sodium hydroxide and sodium carbonate, here described, are hydrolytic.*” Analyses of the solutions at the end points of the reactions, prove that the products of the saponifications are present in the amounts indicated by the theory. “The hydrolytic dissociation of sodium hydrogen carbonate according to the equation: NaHCO,+H.0 = NaOH+H.CO, necessitates a gaseous pressure of carbon dioxide and a continuous loss of the gas with formation of normal sodium carbonate in the solution. A discussion of this question may be more fully entered into a later paper. It is sufficient here to note that the effect due to this cause is very slight. The amount of the hydroxide in this solution is very small. McCoy, Am. Chem. J. (1903), 29, 453, has calculated the concentration to be 2.9X10-. 1” A more detailed discussion will be taken up in a later paper. ~Ann. Chim. Phys. (1844) (38) 10, 327. *Am. Chem. J. (1892), 14, 411. * Secondary reactions take place, to a small extent, not sufficient to affect the accuracy of the method. Some of these, probably due to light rays, are being studied. METHYL SALICYLATE I. 109 TABLE L[V.—The analyses of the solutions described in Tables II and III at the end of the reactions. Methyl alcohol. Salicylic acid. , Solution. fs | Theoret- Theoret- S@nal, Found. aml, Found. Per cent. | Per cent. | Per cent. | Per cent. 0.203 normal NaOH _______ 0.64 0. 67 2.80 2. 66 0.099 normal NaOH ______- 0%324|= aaa 1.37 1.41 0.2 normal NasCO3 —_----_- 0.32 0.314 1.38 1.33 SUMMARY. It is shown that methyl salicylate (synthetic), or oil of gaultheria, when used in foods and drugs, may give rise to the presence of salicylic acid, first, as an impurity in the ester; second, through its hydrolysis. Methods for the separation and quantitative estimation of salicylic acid and methyl salicylate are described. The rate of saponification of methyl salicylate in solutions of sodium hydroxide and carbonate are studied. The work in some of its other phases is being continued. NOTES ON THE SPROUTING COCONUT, ON COPRA, AND ON COCONUT OIL. By Hererert $8. WALKER. (From the Chemical Laboratory, Bureau of Science, Manila, P. I.) CONTENTS. I. EXPERIMENTS ON ENZYMES IN THE COCONUT. II. CHANGES IN THE COMPOSITION OF THE COCONUT WHILE SPROUTING. Ill. Tor AcTION ON CoprRA OF MICROORGANISMS IN PURE CULTURE. IV. THe Propucrion or Free Acid IN CoMMERCIAL Coconut OIL on Lona STANDING. It. EXPERIMENTS ON ENZYMES IN THE COCONUT. The following experiments were made in an endeavor to discover if the coconut, like the castor bean and many other oil seeds, contains a fat-splitting enzyme capable of saponifying outside of the growing nut. COCONUT FOOT. Haperiment [—One hundred grams of the fresh foot in a sprouting coconut were ground with sand and water, and the expressed liquor was strained through cloth. One per cent of toluol was added and the whole allowed to stand on ice over night. (a) Five cubie centimeters of water, 1 of fresh liquor and 0.25 of ethyl butyrate were kept in a water bath at 40° for fifteen minutes and then titrated ; there were required 0.48 cubie centimeters of x potassium hydroxide for neutralization. The mixture was allowed to stand until the next day, when it took 0.12 cubie centimeter more of the,same solution of alkali. (b) Five cubie centimeters of water, 1 of the boiled liquor and 0.25 of ethyl butyrate were placed under the same conditions as the above for fifteen minutes in a water bath; there were required for neutralization 0.45 cubie centimeters of N 6 : ; E 10 potassium hydroxide and on the next day 0.11 cubic centimeter more. (c) Five cubie centimeters of water, 1 of fresh liquor, 0.25 of ethyl butyrate, 1 drop of phenolphthalein, 0.28 cubic centimeter of 3 potassium hydroxide and 0.1 of toluol were placed in a water bath for thirty minutes, then stood =f at room temperature until the next day, when 0.09 cubie centimeters of a potassium hydroxide were required for neutralization. (d) The conditions were the same as in (¢c) with the exception that boiled 111 112 WALKER. N liquor was used. There were required 0.12 cubie centimeter 7 potassium hydroxide for neutralization. Conclusion.—No enzyme capable of hydrolizing ethyl butyrate is pres- ent in the press liquor from the coconut foot. For comparison I give one experiment by Kastle and Loevenhart * working with a 10 per cent extract obtained from the pancreas of a pig. One cubic centimeter extract, 4 of water, 0.26 of ethyl butyrate and 0.1 of toluol were kept at 40° for fifteen minutes and showed an increase in acidity corresponding to 1.63 cubic centimeter T0 potassium hydroxide. A similar test with the boiled extract showed no increase in acidity whatsoever. My Lxperiment I was continued as follows: (e) Eight cubie centimeters of fresh liquor, 5 of coconut oil and 0.1 of toluol Hight cubic centimeters of = potassium were allowed to stand one week. hydroxide were required for neutralization. (f) Hight cubic centimeters of boiled liquor, 5 of coconut oil, and 0.1 of toluol werd allowed to stand one week. 19.9 cubic centimeters 2 potassium hydroxide were required for neutralization. The boiled liquor showed a considerably greater acidity on standing than did the fresh, hence it is evident that no hydrolysis by enzymes had thus far been proved. The nut from which this foot was taken was perfectly sound and free from mold, but the inner surface of the meat next to the foot had begun to soften and had a greasy feel. A portion of this softened meat was dried and expressed, yielding an oil containing 3.3 per cent of free fatty acids, showing that hydrolysis to a marked extent had taken place in the growing nut. COCONUT MEAT. Experiment II—The sprouting nut used for this series contained a foot which almost filled it. The meat remaining was ground in a sausage grinder and a cream-like emulsion pressed out. (a) The action with ethyl butyrate-——The conditions were; 5 cubic centi- meters of water, 0.25 of ethyl butyrate, 1 of toluol and 1 of cream, with the following result : | Conditions. 14 hours at Neutralized, let | Let stand 1 day stand at room more at room Not neutralized, room temper- 40° C. temperature tempera- ature for | ] day. ture. 2 days. N Ne N N Fre: res | 0.45 =-K 2:ce. —K a KG . 92 ec. — K HMreshicren ees ae—s=s— 0.45 ee. 10 KOH | 0.12 ee 10 KOH | 0.10 ce i0 KOH | 0.92 ce 10 KOH Boiled cream __----- Nee 35 ce. — KO 0.35 ec. 5 KOH = | N N PLOICCH=— b -— K 0.10 ce 10 0.10 ee po KOH gh IN 0.45 ce. 75 KOH Amer. Chem. Journ. (1900) 24, 491. NOTES ON COCONUT, COPRA, AND COCONUT OIL. 1138 (b) The eream alone with 1 per cent of toluol as an antiseptic was allowed to stand four days in the incubator. (1) Five cubic centimeters of boiled cream required 7 cubie centimeters of a potassium hydroxide to neutralize. (2) Five cubic centimeters of fresh cream required 12.8 cubic centimeters of N : F , : — potassium hydroxide to neutralize. (c) After pressing out the cream used in (a) and (0), the residue was ground with sand and water and two fractions pressed out in the hydraulic press: A (up to 250 kilograms per square centimeter) and B 250 to 450 kilograms). These samples were kept under the same conditions as in (b) for four days and aliquot portions, each of 5 cubic centimeters, were titrated with x potassium hydroxide. The following is the result: A, boiled and fresh, and B, boiled and fresh, each took 0.05 cubic centimeter of PE potassium hydroxide to neutralize. The cream from the first pressing (a) is therefore the only one showing any indication of enzyme activity. (d) Five cubie centimeters of the cream from the first pressing (a), 5 of coconut oil and 2 of toluol were placed in an incubator for four days; there were required 12.4 cubie centimeters of x potassium hydroxide for neutraliza- tion. A control made under the same conditions with boiled cream required 6.6 cubic centimeters. The greater increase in acidity of the unboiled cream seemed at first to indicate enzyme action; but plate cultures made from the two tubes showed a considerable number of mold and bacterial colonies, these being more numerous in the unboiled than in the boiled cream. Therefore, it seems more reasonable to attribute the increase in acidity to the inefficiency of the antiseptic used, rather than to a specific enzyme action. COCONUT MILK. Bxperiment I1I—The nut used had begun to sprout, its inner space being almost completely filled by the endosperm. About 50 cubie centimeters of milk were obtained and tested, the conditions being as follows: 5 cubic centimeters of water, 0.25 of ethyl butyrate, 1 of toluol and 1 of milk. ae ee 15 minutes at 12 hours at | Conditions. 40°. 35-100. | . N N Whoo .28 ce. — KOH | 0.35cee.— K | | boile 0 ee 10 KOH 39 CC 10 KOH é N N TUG 0.28 ce. KOH | 0.38 ec. ~ K | ce 10 KOH | 0.38 ce 10 KOH The results show no evidence of the presence of an enzyme in coconut milk. 114 WALKER. MEAT AND MILK. Haperiment IV.—(a) The expressed cream from the meat of a nut just beginning to sprout was used, the conditions being as follows: Five cubic centimeters of water, 0.25 of ethyl butyrate, 1 of toluol and 1 of cream. ~ Not neutralized. Neutral- Pre ized, and Incubator 24 Conditions. At start. let staal Room Haare 24 hours, | tempera- ture 24 | hours. Be Be | ee. ee. ce. ce. N nboiled 222332 0.47 ce. 10 KOH | 0.18 1.82 1,24 2.00 Boiled | vides eee ee _| 0.47 ce. 0.16 0.70 0.60 | 0.47 (b) Five cubic centimeters of water, 0.25 of ethyl butyrate, 1 of 1 to 1,000 formalin and 1 of eream were used. The unboiled mixture after twenty-four Tas 5 - e N : 5 hours at room temperature took 0.55 cubic centimeter of 10 potassium hydroxide and the boiled mixture 0.48 to neutralize. Here again, although the unboiled cream increases in acidity fairly rapidly in the presence of toluol, the increase must be due chiefly to inefficient antisepsis, as it is almost entirely inhibited by formalin in a dilution of 1 to 7,000. Experiment has shown that formalin of this strength has practically no action on enzymes. Experiment V.—The cream used in Laperiment IV was treated with alcohol and ether, the precipitate washed with alcohol and finally with ether and then dried in a vacuum over sulphuric acid. One gram of this powder rubbed up with 20 cubic centimeters of water yielded a milky liquor which was tested for the presence of enzymes, the following being used: Five cubic centimeters of water, 0.25 of ethyl butyrate, 1 of toluol and 1 of the liquor, the mixture being kept in the incubator at 35° to 40° for two days. N Both the boiled and unboiled liquor required 0.8 cubie centimeter of i0 potassium hydroxide to neutralize. It is evident that the precipitate obtained from coconut cream by the above method contains no fat-splitting enzymes. COCONUT FOOT AND COCONUT OIL. Experiment VI.—(a) One hundred grams of foot from two nuts with sprouts about 1 meter long were mixed in a mortar with 100 cubie centimeters of fresh coconut oil, 100 of water and 1 of chloral. The mixture was well ground for about one hour to prepare a good emulsion, it was then strained through cloth and 25 cubic centimeter portions were placed in small, stoppered Erlenmeyer NOTES ON COCONUT, COPRA, AND COCONUT OIL. 115 flasks. The samples were then titrated at various intervals with N potassium hydroxide, the result being given in cubic centimeters required to neutralize: Cubic Time. centimeters. At once 13.5 1 day 12.0 2 days 12.2 7 days : 13.8 8 days 18.5 Cultures made at the end of the last period showed the presence of molds and bacteria. (b) One hundred grams of meat from the same nuts as were used in (a) were ground with 75 cubic centimeters of fresh coconut oil, 75 of water and 1 of chloral; 25 cubic centimeters of the strained emulsion required the following amounts of * potassium hydroxide for neutralization: Cubic Time. centimeters. At once 10.0 1 day 11.0 2 days 11.5 7 days 24.0 8 days 27.5 Cultures made at the end of eight days showed that the flasks were no longer sterile; they had all become slightly discolored at this time and im some cases possessed a bad odor. The slight increase in acidity in this comparatively long term must therefore be attributed to the actions of organisms, not to enzymes. ATTEMPTS TO RENDER AN ENZYME ACTIVE. Connstein, Hoyer and Wartenberg,? working with lipase from the castor bean, have shown that its rapidity of action is greatly increased by the addition of a small amount of acid. The following experiments were made to discover if this might not likewise render active any enzyme of the coconut which might be present. COCONUT FOOT AND COCONUT OIL WITH HYDROCHLORIC ACID. Experiment VII.—(a) Fifty grams of the foot from a nut with sprout about 1 meter long were shaved into fine pieces and ground in a mortar, first with 2 grams of chloral hydrate, then with 50 grams of fresh coconut oil and finally with 25 cubic centimeters of A hydrochloric acid. Ten cubie centimeter portions of * Ber. d. Chem. Ges. (1902), 35, 3988. 116 WALKER. the resulting mixture were then placed in small, stoppered Erlenmeyer flasks and N titrated from time to time with iG potassium hydroxide. Cubic Time. centimeters. At once 10.0 1 day 10.0 3 days 11.5 5 days 10.5 8 days 11.0 14 days 11.3 23 days 11.0 The result shows that practically no hydrolysis had taken place in twenty-three days. The flasks at the end of this time remained sterile. COCONUT MEAT WITH HYDROCHLORIC ACID. (6) One hundred grams of meat from a nut with a sprout about 75 centi- meters long were thoroughly ground with 2 grams of chloral hydrate and 50 cubic centimeters of 5 hydrochloric acid; 10-gram samples of the mixture were placed in small, stoppered Erlenmeyer flasks and titrated with * potassium hydroxide after the addition of absolute alcohol and phenolphthalein. The result was as follows: Cubic Time. centimeters. At once 8.5 2 days 10.0 4 days 9.1 ' 13 days 11.1 19 days 11.2 22 days 11.5 The flasks were practically sterile at the end of twenty-two days. The slight apparent increase in acidity may almost be accounted for by experimental error, as in a non-homogeneous mixture such as this an exact titration is exceedingly difficult. EXPRESSED LIQUOR FROM FOOT AND MEAT OF A SPROUTING NUT IN EMULSION WITH COCONUT OIL AND GUM ACACIA. ‘ Experiment VIIJ.—One hundred grams of foot from a sprouting nut, 100 of meat from the same nut, 100 of sand and 100 cubie centimeters of a 2 per cent chloral hydrate solution were ground together in a mortar for two hours and the whole mass expressed; there resulted a muddy liquor consisting of water and partially emulsified oil. Fifty grams of coconut oil were rubbed up with 20 grams of gum acacia and 100 grams of this press liquor to form a permanent emulsion, 10 cubic centimeter samples of which were transferred to flasks and NOTES ON COCONUT, COPRA, AND COCONUT OIL. JULY N titrated with 0 potassium hydroxide as in the previous experiments. The result was as follows: | ] | Conditions. At once.| 2 days. | 6 days. ce. ce. cc. | Press liquor-+-oil 7.0 6.7 982 Press liquor alone Gh 9.5 12.3 No cultures were made from these flasks. CREAM EXPRESSED FROM COCONUT MEAT, WITH ACETIC ACID. Experiment IX.—(a) The entire meat from a nut with a sprout about 30 centimeters long was ground in a meat grinder and pressed in a hydraulic press up to 450 atmospheres, 50 cubic centimeters of the thick, cream-like press liquor, 25 cubic centimeters of a 1 per cent chloral hydrate and 2 per cent acetic acid solution were ground to an emulsion, weighed in separate portions of 10 grams each and titrated in alcoholic solution with x potassium hydroxide. Cubic Time. : centimeters. At once ey 1 day 37 3 days 1.9 6 days 2.05 31 days 2.15 : PRESS CAKE FROM MBAT, WITH COCONUT OIL AND ACETIC ACID. (b) The press cake from (a) was dried over sulphuric acid at room tem- perature and ground to a fine powder; 2.5 grams of this powder with 22.7 of coconut oil and 10 of acid chloral solution were ground together for two hours in a a mortar; 10-gram samples were taken for titration with a potassium hydroxide with the following results: Cubic Time. centimeters. At once 1.10 2 days 1.08 7 days 1.08 (c) A parallel experiment with the dried press cake from castor beans was made at the same time; 2.5 grams of the cake, 22.7 of coconut oil and 10 of acid chloral solution being ground together, and 10-gram portions were titrated with S potassium hydroxide, with the result recorded below: Cubic Time. centimeters. At once 2.7 1 day 12.4 2 days 15.0 118 WALKER. EXTRACTED, DRIED AND PULVERIZED MEAT WITH COCONUT OIL AND ACETIC ACID. Baperiment X.—(a) The meat from a nut with a sprout about 30 centimeters long was cut into fine pieces and extracted with cold ether until practically all the oil was removed; the remainder was then dried in a current of air at room temperature, and finally in a vacuum over fused calcium chloride. When dry, it was ground to a fine powder and again dried over calcium chloride; 2.5 grams of this powder, 22.7 of oil and 10 of an acid chloral solution were ground together and 10-gram portions titrated with < potassium hydroxide, with the following result: Cubic Time. centimeters. At once 1.18 2 days 1.15 7 days f 1.05 (b) The foot from same nut was extracted with cold ether, dried in a vacuum, and ground to a fine powder; 2.5 grams of this powder, 22.7 of oil and 10 cubic centimeters of acid-chloral solution were ground together and 10- = gram portions used for titration with x potassium hydroxide. There was practically no change, as is shown below: Cubic Time. centimeters. At once 1.38 2 days 1.35 7 days 1.32 SUMMARY. Experiments were first made with coconut meat, milk and the foot under varying conditions, to determine if they were capable of hydroliz- ing ethyl butyrate according to the methods employed by Kastle and Loevenhart, toluol being used as an antiseptic. In most cases no increase in acidity was noted. However, when hydrolysis did occur the tubes were found not to, be sterile, the conclusion being that toluol is not a sufficiently strong antiseptic to prevent the growth of organisms in coconut extracts. When formalin was used in a dilution of 1 to 7,000 the hydrolysis was entirely stopped. Formalin of this strength has almost no effect on_ enzyme action. After this failure to hydrolize ethyl butyrate by means of extracts from the meat, milk and foot, attempts were made to emulsify and hydrolize pure coconut oil according to the method of Constein, Hoyer and Wartenberg, using different portions of the coconut, instead of castor beans. In no case were these attempts successful under sterile conditions, although parallel experiments carried on with castor bean press cake showed a decided hydrolysis. In none of these experiments, which extended over a period of seven months, is the slightest proof given of the existence in the coconut of an enzyme capable of hydrolizing fat outside of the growing nut. The NOTES ON COCONUT, COPRA, AND COCONUT OIL. 119 cause for the destruction of the fat in the growing nut must therefore be sought elsewhere. A discussion of the changes taking place in the sprouting nut is given in the following chapter. II. CHANGES IN THE COMPOSITION OF THE COCONUT WHILE SPROUTING. Four pairs of sprouting coconuts of different ages but approximately of the same size were selected for this work and their composition de- termined as follows. TOTAL WEIGHTS. After measuring the length of the sprout, the total weights of the whole nuts (including the shell but free from husk), the milk, foot, meat and sprout (with roots) were determined at once. MEAT. Three samples of 10 grams each were taken from each nut for analysis, viz: Ten grams from that portion of the meat nearest the foot; Ten grams from that portion of the meat farthest from the foot; Ten grams as an average sample of the remainder which in Table I is calculated on the total. Moistwre.—Vhe materials were dried for five hours at 100° C. Oil.—The dried meat was ground to a fine pulp in a mortar and extracted in a Soxhlet cone with chloroform. Sugar.—After removal of the oil the remainder was extracted with 50 cubic centimeters of water for three hours in the same apparatus as was used for oil extraction, the solution was then placed in a 100 cubic centimeter flask, clarified with basic lead acetate, the excess of lead removed with potassium oxalate, the whole diluted with water to 100 cubic centimeters, filtered, and after inversion, the sugar determined in 25 cubic centimeters of the filtrate by Fehling’s gravimetric method. Crude fiber.—This was determined in the residue from the extraction with water, according to the method of the Association of Official Agricultural Chemists. = FOOT. Sugar.—Ten-gram samples were ground to a fine pulp and placed in a 100 cubic centimeter flask with about 50 cubic centimeters of water, they were then allowed to stand at room temperature with occasional shaking for three or four hours and diluted to the mark after the addition of basic lead acetate and potassium oxalate. Twenty-five cubic centimeters of the filtrate were inverted and the sugar determined by Fehling’s gravimetric method. Orude fiber.—Determined according to method of the Association of Official Agricultural Chemists. MILK. Sugar.—Fifty-gram samples were clarified, diluted to 100 cubie centimeters and 10 cubic centimeters taken for the Fehling determination (equivalent to 5-gram samples). FREE FATTY ACID IN THE OIL FROM THE MEAT. After determining the percentage of oil in the meat, the oil was titrated in T alcoholic solution with = potassium hydroxide, phenolphthalein being used as an indicator. The results are given in the following table: WALKER. 120 O'SE | LLG L6 TU LG 16 8 FP Or #9 Ts B9 LG 16 97% T's Ti E Oy ear dec ae eee a yooy ul wesng Cis 8's LG 1b) “qSO[ UOT] BULL} 9(T 9°0 £0 10 £0 60 0 Tay GhOls 4: | aaah oe Seabee oe yoo} UL Jaqy apniy (ne 6% 6T o@ TL 61 8° 0G LL 8'T G9 9% Z 01 ind TS TT | ticseany io oo ie DOUBLE, pee 6°T | 6L°0 LT L4L0 L1°0 LT LL'0 9'T OTi0) LE LT0 9% 92 '0 1l'% 120 ~q003 a} 1soddo survis OT ur seqy apnap *qOOJ LB9U YVOUT ON #a'0 CoG a) 6°T 610 Lt 110 Ts Ts 0 ara OW = |r }00J LBOU SULBIS OT = @T | SOT qT 0% (it) 97% 6°L 8°s PPL 67% 9°OL L5§ : 8 °FL IP (OS) Ne [bee = ceo er dOUBL VE 80 | 800 gt | &T0 960 ge | 980 GF | or0 ep | sO einen nea pe | FSO [9007 aytsoddo survis nf uransng “JOO IBOU JBI ON 80°0 ¢0 ¢0 0 oT cr'0 8°0 80°0 tam &10 0G (EW 9 S JOO} 1BAU SUIBIS OT coq | S'ss 8°LP 169 Lot | CLP TOL | 8°3¢ ‘0G | G's 061 | L°8F C6L | 8'8h SO iliea| wtaser aay, aan aoUBlB_ B'°9G | G9'S GTS | ore cor FCS | POS 8°8¢ | 88°¢ e1¢ | e4'¢" G%g | cae L°GG | LoS ~qooy oy 1soddo stuwis vf Ul LOJRM “WOOF ABOU FBIM ON OL's ech | Sop 4 ‘6P G6 OF STF | 8L'F PGP POP Oro | ORS | qOOJ LBIU SUIBIT OT ¥ 9S GL PLE P'6P COL | 8°Le ‘CIT | 418 TOL | G's ‘OIT =| L°S8 “GST | 6°GE YEG Tineg| eee setae aL aouVv[ vd Lag | Lg-¢ GFE | GP'S 69'S L118) | -L0-& G94 | G9 T°LG | TLS 9°6% | 96°G 91g | 9L'é ~yooy oy 1soddo suvis of a ur TIO “JOOJ LBOU J BIT ON 64h 8°68 | 86° 9-98 | 99°& Leh | Leh 6°Uh | 6L'P (On CS ien | MOG Sie | ane JOO} IBAU SURI OT 4ue? Sauron | Yor | tases | Voy | TUSIOM vo | Us ean | og. | TUSIOM 292 |quaiom | Wag | TSIM | Voy | IUBIOM 6&8 FFG OL or ‘paystoa JON | “‘paysfoa JON | 0 (One eninioiines eine sjoor puw ynords fo yy SIO MA SFL rane OOF LOS 0F 68S IGP NG Pores se hie cies | feet VBIUL JO SULBIS UL IYSIOM [BIOL S1a 826. SLL 06 6F Gr i 6G GI ie ee eee JOO} JO SUIBAS UL ISOM [BIOT, 0 0 Oot LLT 006 GLs 408 PLES ee | ee Y[LUL JO SUIBIS UL IY SIOM [BIOL O9L G18 200 ges 098 C68 FOOL SEO Le bate e etal Ca Seen ebiaa an ips eae lel ({eqs puv sjoor Ynoids Surpnyour) ysny UWLOLJ DAIT JNU JO SUVA Ul JYSIOM [BIOL 0 86 &6 6e ge 6 ba { --paqnords 10N Ee ee ao eae aie siojotaI}zU9O Ur ynords Jo yySsue'T 8 oh 9 ¢ G iTS “TOqQUInNN “‘Buynosds Burunp ynwos0o ay, Ur aonid bun, sabunyg—] @IavL NOTES ON COCONUT, COPRA, AND COCONUT OIL. iA DISCUSSION OF THE CHANGES TAKING PLACE. Miulk.—Vhe total quantity of milk shows a marked diminution from 374 grams in an unsprouted nut to nothing when the sprouts have at- tained a height of 93 centimeters. A decided loss in the sugar content of the milk takes place at the same time, as this constituent falls from 2 per cent and 2.3 per cent in the milk from the unsprouted nuts to 0.3 per cent in the ones which have sprouts 38 centimeters in height. Meat.—Here also a decided loss in total weight is evident, as it drops from 457 grams in nut number 1 to 148 grams in number 8. The loss seems to be due to a direct absorption by the foot, the process taking place at first only in that portion of the meat located near the latter, but increasing rapidly as the endosperm grows larger and comes in contact with the entire inner surface of the nut. Oil in the meat—The loss in total weight of oil is fairly proportional to the loss in total weight of meat, the percentage of oil in the meat remaining constant within the somewhat wide limits of individual varia- tion. However, during the early stages of germination there is apparent a certain concentration of oil near the foot, with a corresponding loss in that portion of the meat farthest away. Water in the meat.—As is the case with all the other parts of the nut, the meat gradually loses water by evaporation through the shell and sprout during the process of germination. Sugar in the meat—TVhe percentage of sugar decreases from 4.1 per cent in the unsprouted nut (number 1) to 1.2 per cent im number 8. The loss is probably due to the absorption of sugar by the foot, as in all cases there is considerably less sugar in that portion of the nut in direct contact with the endosperm than there is in the parts farthest away from it. Crude fiber in the meat.—No decided change in the proportion of this constituent can be observed. It is absorbed at practically the same rate as the rest of the meat. FOOT. Total weight.—Vhe total weight increases from 19 grams in number 1 to 228 grams in number 7. Sugar in the foot.—There is apparently a loss in the percentage of sugar (although not in its total weight) until the foot completely fills the nut, at which time there is a rapid gain. This phenomenon is probably due to the fact that the foot at first draws its sugar chiefly from the milk by which it is almost entirely surrounded. However, as it continues to grow, it soon exhausts the sugar in the milk, and only when it has completely filled the nut and come into intimate contact with the inner surface of the meat has it an opportunity to continue the process of sugar absorption and also one of sugar creation, possibly from the oil, or possibly from oil and crude fiber. 122 WALKER. Crude fiber in the foot.—A slight imerease in this constituent (from 1.1 per cent and 0.9 per cent in numbers 1 and 2, to 2 per cent and 1.8 per cent in numbers 7 and 8 respectively) will be noted. Tf we consider the total weights of the constituents, oil, sugar, and fiber, the following changes may be remarked : Oil—A decrease from 163 grams in number 1 to 78 grams in number 8 is observed, or a total loss of 85 grams. Sugar.—There appears to be a considerable loss in total weight of sugar during the intermediate period of germination, which is however again made up at the time the foot fills the entire nut. Increase in sugar takes place in the foot. Considering the first and last of the series, numbers 1 and 8, the following changes in sugar content have taken place: Portion of the nut. Loss. Crude fiber—There is a slight loss in the total weight of fiber in - foot and meat, which is more than made up by the increased weight of sprout and roots. Free fatty acids and oil from the meat.—F rom the beginning, the oil from the meat nearest the foot is invariably richest in fatty acids, number 1, for instance, yielding 0.92 per cent in that portion, while the balance contains but 0.27 per cent. This difference becomes more marked as germination progresses; it is only when the foot has come in complete contract with all the meat that an increase in fatty acids throughout the whole nut is observed, indicating that oil, to be in a condition for absorption, must be hydrolized. This hydrolysis may take place as the result of an enzyme in the foot, or be caused by one in the meat, which is dormant until rendered active by some product of metabolism in the foot. However, as I have stated in the previous chapter, it was not possible to prove by an increase in free acid the presence of any fat splitting enzyme in the coconut. Such an enzyme may exist, but under such conditions that any large excess of free acid must be used up by the growing plant before the process can continue. A summary of the changes to be detected by chemical analysis of the growing coconut is as follows: Oil is lost by the meat; it is not taken up as such by any other portion of the nut, but is either burned to furnish energy for the growing NOTES ON COCONUT, COPRA, AND COCONUT OIL. UB} plant or is split up, being transformed by progressive synthesis into sugar and finally to cellulose. Sugar is lost by meat and milk, but a corresponding quantity is gained by the foot, the total amount in the nut remaining approximately the same. A small amount of crude fiber is lost by the meat, but a much larger quantity is produced in the sprout and roots. Ill. THE ACTION ON COPRA OF MICROORGANISMS IN PURE CULTURE. It has been shown previously * that moist copra is readily attacked by microdrganisms with consequent splitting up and destruction of the oil and it has also been proved that the action of such organisms is most. pronounced when the copra has a water content of from 10 to 15 per cent. With this content of moisture the mold growth largely predomi- nates over that of the bacteria. When much more water is present, and the bacteria are in excess of the molds, destruction of fat is greatly diminished. These observations led logically to the belief that hydrolysis of oil in copra was due to the action of molds alone, although the data available at the time the previous work was done did not exclude the possibility of symbiosis and interdependence in this fat-splitting process between molds and bacteria. Dr. Edwards, of the Biological Laboratory of this Bureau, undertook further work to settle this question definitely and in pursuing it separated as many different organisms as possible, some fifteen in all, from several samples of moldy copra and coconut meat, finally succeeding in isolating in pure cultures the majority of the erowths present. As a culture medium he used sterilized coconut meat in most instances. The subsequent procedure was as follows: Ten-gram samples of anhydrous copra were placed in large test tubes stop- pered with cotton, and after the addition of 1.50 grams of water these were sterilized in an autoclave for half an hour. It was found by experiment that about 0.10 gram of water was taken up by the copra during sterilization, so that the samples thus prepared contained approximately 13.8 per cent of moisture, an amount which had been found previously to favor the growth of both molds and bacteria. The tubes were inoculated after sterilization with pure cultures of the organisms previously isolated, and allowed to stand at laboratory tempera- ture (26° to 30°) for forty-one days. Cultures were then made from each tube. All tubes were dried to constant weight at 100° in order to determine the change in weight of the dry copra. The latter was next extracted with chloro- form to determine oil and finally with hot water to take out the sugar, which was inverted and determined with Fehling’s solution. The changes taking place in forty-one days are shown in the following table. Only those tubes which showed a good growth in this time and were proved by cultures to contain only one organism are noted. ®This Journal (1906), 1, 123. 124 WALKER. Tas_e II1.—The decomposition of copra by bacteria and molds. Te Gi, sugar, | Undeter S| 22 OF al orale me ral fees firey alae Peetu fae Per ee ly 3 PY a oc 7 bal a a Re 7] ® i go) it I ® oo |g Shes | =) Ba a ey 3 g me a 28S) 3 < This figure represents the clear portion of the oil. After shaking the bottle to obtain a representative sample together with sediment, ete., a figure of 7.6 per cent was obtained. NOTES ON COCONUT, COPRA, AND COCONUT OIL. WAT was being continued in the oil, either by portions of the mold carried over through the factory filters, or by enzymes which were expressed and which would find their way into the final product. In order to determine if this supposition were correct, a number of 50-gram samples of the oil were subjected to different treatments, such as filtration to remove sediment and most of the water, the addition of antiseptic, sterilization by heat or by a combination of all three of the above processes. The oils were heated at a temperature of 100° in a water oven for two hours; ordinary quantitative filter paper was used for filtration. Two samples from each treatment were prepared, one being kept in a 100 cubic centimeter bottle half full, the other in a 50 cubic centimeter bottle filled to the neck. Unless otherwise stated, samples were sealed with paraffin and kept in the light. One sample, in the case of numbers 7 and 8, was kept in the light in 100 cubic centi- meter Erlenmeyer flasks with sterilized cotton plugs, where air and light might be expected to play the leading part in any change produced; the other sample was kept in a wooden box covered with black paper. The following table shows the change in acidity during a period of two years: ; TABLE III.—Change in acidity of coconut oil standing under different conditions for two years. 0 Pe (oe tee (ae | 2 | ge. 22 lee | 88 je? | Se ie ‘ Description of |ESE| bE |Go | BS |geag| os No. Description of oil. pecuice and ae a a | es 5 2 a es S| cel i condition. o28 ae Hmog os 5 ae |2eea) 2a | gre) SS |gno| ot | e & |= ~ = a | 1 | From original bottle unheated, |ja. Small bottle__ 7163 | de) 8.6 | 11.6 0.4} 4.0 unfiltered, no antiseptic. te. Large bottle-_| 7.6} 11.0 3.4] 13.4 2.4) 5.8 | 2 | Unheated, filtered, no antiseptic _ a Small bottle} 7.4 Bis Hots bod O09 |) 28 | b. Large bottle__ 7.4 8.8 1.4) 11.2 2.4] 3.8 3 | Unheated, unfiltered, +0.05 per \ja. Small bottle__| 7.6 9.6 2.0) 10.9 IB |) 8483 cent chloral. to Large bottle__| 7.6 9.2 1o@G| aal@) 2.7) 4.3 4 | Unheated, filtered, +-0.05 per cent \ja. Small bottle__| 7.4 8.7 1.3 9.5) 0.8] 2.1 chloral. ty Largebottle.| 7.4) 87] 1.3] 102] 1.5] 2.8 5 | Heated, unfiltered, +0.05 per \ja. Small bottle__| 7.6 8.9 683 9.9 I O5)) 253 cent chloral. tb. Large bottle__| 7.6 9.0 1.4) 11.2 PN BS 6 | Heated, filtered, +-0.05 per cent |ja. Small bottle__| 7.4] 8.2 0.8 8.8 0.6) 1.4 chloral. \b. Large bottle__| 7.4 8.9 1.5) 10.4 1.5] 3.0 7 | Heated, unfiltered, +0.05 per | cent chloral, kept in Ehrlen- |{®: 12 ight ---- 7 i 10 e210) ae Sd ee meyer flask with cotton stop- \b. In dark _____- toe eb 20) Cabs Be per. | | 8 | Heated, filtered, +0.05 per cent fa. Inlight === oes |) an@ |) -8bG |) 31,0) Wl] % chloral, keptsameasnumber7. |\b. In dark _____- 7.4) 11.6 4,2 | 915.9 4.3) 8.5 a Since no increase in acidity due to the action of light could be observed after one year, the samples previously kept in the dark were taken out and placed alongside the others serving simply as a check on the determinations. 128 WALKER. It is necessary for the interpretation of these results to consider the many factors which may enter into the decomposition of a freshly pre- pared, commercial coconut oil. First, we may have present fat-splitting molds, albuminoids, sugar and water, which cause the turbid appearance of commercial oils. It has been shown in a previous paper® that mold action on copra is’ the principal factor in determining the initial acidity of an oil, and that . these same molds in the presence of sufficient nutritive matter may effect the rapid decomposition of even a pure oil. Second, soluble or insoluble enzymes produced by these molds may be the cause of the rise in free fatty acid. Third, surface oxidation by the air, possibly assisted by light, may take part in the decomposition. ‘This surface oxidation is always accompanied by a pungent, disagreeable odor, and the formation of aldehydes and peroxides. Houwrth, simple hydrolysis by heat and moisture may be a factor. Considering first sample number 1 of this series, it is evident that 1-b, the sample in the large bottle, may be subject to any one or all of the foregoing factors causing increase of acidity, while 1-a, being in a prac- tically full bottle, may be affected by any influence except that of oxida- tion by the air. Any marked increase in acidity, then, of 1-b over 1-a would be due to surface oxidation. By reference to the table we find that during the first year there has been practically no difference in the rate of increase of acidity between 1-a and 1-b, which show respectively a gain of 3.6 and 3.4 per cent. The second year, however, indicates a decidedly different condition. While 1-a has only gained 0.4 per cent free acid, 1-b has inereased 2.4 per cent, leaving a difference of 2 per cent which can only be due, aside from experimental error and an individual variation in the samples, that should at most amount to no more than a few tenths of a per cent, to surface oxidation by the air. The question naturally arises, why should a period of two years be required for this difference to show itself? This can be accounted for by two theories: First, the presence of molds and of nutritive bodies such as sugar and albumen, which would be in the oil in larger quantity in the first year than in the second, is not favorable to the formation of peroxides of the fatty acids, a fact I have previously noted in testing pure and com- mercial oils for peroxides and aldehydes ;* second, it is possible that this inhibiting effect may be mutual, so that oxidation once started would tend to kill off or check mold or enzyme activity in 1-b sooner than this activity would naturally cease through lack of nutriment in I-a. In either case, the result would be the same and the combined effect of the °This Journal (1906), 1, 139. 7 Loe: cit., 139. NOTES ON COCONUT, COPRA, AND COCONUT OIL. A) two processes, each ending ultimately in the production of free acid, might be less than either of them working singly. ‘The probabilities are that oxidation does not set in until most of the nutritive substances present in the oil are used up, thus it would be natural to expect that for a certain period of time no difference in the rate of acidification of an oil, due to the size of the container, would be observed. Sample number 3 was. filtered, removing molds, albuminous matter and any enzymes insoluble in oil, together with most of the water. Dur- ing the first year an increase in acidity of 1.4 per cent was noted. The difference between this figure and 3.6 per cent, the increase of 1-a for a corresponding time, gives 2.2 per cent which may be attributed to molds and insoluble enzymes. Practically no difference, due to size of bottle, is observed in number 2 as well as in nearly all the other samples, during the first year, although it is quite marked at the end of the second year. Number 3 differs from number 1 only in containing a small amount of antiseptic. It shows 1.6 per cent less increase during the first year than number 1. This can be due only to inhibition of molds. During the second year 3-a has increased 0.9 per cent more than 1-a, a fact for which I can find no explanation, except that during the second year some surface oxidation may have taken place, even in the small bottles. Five cubic centimeters of oil had been removed for titration at the end of the first year, thus leaving a small air space. This being the case, somewhat wider variation in acidity might be expected. The figure of 0.4 increase for the second year in the case of number 1-a seems exceptionally low, compared with the other samples during this period. 3-a has increased only a trifle more in acidity as compared with 2-a, thus proving that the addition of antiseptic has about the same effect as filtration and that most of the difference in behavior of a filtered and an unfiltered oil is due to the removal of molds and insoluble enzymes. Number 4 was filtered and treated with antiseptic with results which practically correspond with those of 2 and 3. Filtration appears to be slightly more efficient than adding antiseptic. It is quite possible that chloral in the strength used does not have an immediately fatal effect on fat splitting molds, although it certainly inhibits their action to a. very marked extent. Numbers 5 and 6 have both been heated at 100° and treated with antiseptic, thus eliminating mold and enzyme action, and in the case of full bottles leaving only the factor of hydrolysis to be considered. As would naturally be expected, the filtered sample, 6-a, has increased in total acidity considerably less than 5-a, from which the water was not removed. Only number 6 of the whole series shows practically the same difference at the end of each year between the full and the half full 130 WALKER. bottles. The slight, but regular, increase in acidity of 6-a is probably due to the hydrolytic action of a trace of water not removed by filtra- tion, combined with the free fatty acids already present. Some slight oxidation also may have taken place, as the bottle was not absolutely free from air. : Numbers 7 and 8, sterilized and kept under antiseptic conditions in Erlenmeyer flasks, with a large oil surface exposed and free access of air through the cotton plugs, practically doubled their percentage of free acid in two years. The filtered samples of number 8 show considerably more increase than the unfiltered oils numbered 7, a fact which tends to confirm my belief that as a general rule the freer an oil is from moisture and impurities the more quickly is it subject to oxidation when exposed to the air. The samples kept in the light have not increased in acidity any more than those kept in the dark, in fact, 8-b at the end of one year contains 0.6 per cent more free acid than 8-a. In view of this latter work, indicating the considerable variation in acid content of the same oil which may be brought about by different sized containers with consequently varying amounts of oil exposed to the air, it was decided to bring to a close the series of oils described in a previous paper ® and which were set aside to determine the amount of free acid which might be produced on standing, since after the first year, most of the further increase in acidity would be dependent to a very large extent upon the amount of oil surface exposed to air in the bottles. No attempt was made at the time when these samples were prepared to exclude oxidation by keeping them in filled bottles, and with few ex- ceptions no record was kept of the quantity of oil in a bottle, so that wide variations in acidity after the first year were to be expected. For convenience, a description of the oils as first made is reprinted. here. together with a table giving their increase in free fatty acid from the time of their preparation up to the present date. DESCRIPTION OF OILS USED AND DESCRIBED IN TABLE IY. (A) Expressed oil from vacuum-dried copra. Has been heated for two hours at 100° and filtered twice through paper. A light-colored, clear oil with the characteristic coconut taste and odor. (B) An oil similar in every respect to “A” except that it was prepared from copra dried at 80° to 90°, without vacuum. (1) Fresh coconut meat grated and dried at 80° to 90° on August 16, 1904; was allowed to stand in a covered specimen jar until March 11, 1905. At that time it was still of a pleasant odor and taste, although both odor and taste were not quite as good as when the specimen was freshly prepared. No mold ®Ibid., 118. NOTES ON COCONUT, COPRA, AND COCONUT OIL. 131 growth was present. A sample of oil was expressed from a portion of this copra by using a hydraulic press with a final pressure-of 450 kilograms per square centimeter. This oil, after filtration, was of a light yellow color and it was of a pleasant, although slightly burnt, odor and taste. (2) Oil number 1 was heated at 100° for three hours, while at the same time a current of air in a partial vacuum was passed through it. This process leaves the color and free acid unchanged, but removes almost all of the burnt odor, leaving a bland, almost tasteless, oil. (3) An oil from the same copra as numbers 1 and 2, but prepared by extraction with petroleum ether. Afterwards it was treated in the same manner as number 2. It differs from numbers 1 and 2 in being practically colorless. (4) Commercial coconut oil treated with alcohol and animal charcoal and then filtered; the aleohol was afterwards distilled and recovered. This oil was rather unpleasant to the taste, but it had no odor. (5) Commercial coconut oil treated with live steam; this removes the odor, but the unpleasant taste remains. (6) Fresh meat, ground and dried in vacuum at 70° to 80°. The oil was expressed and once filtered; it possessed a very pleasant, coconut-like odor and taste. It still contained a considerable amount of sediment. (7) Coconuts cut in halves and dried in vacuum at 75° to 85°. The oil expressed and filtered twice. It had a very pleasant odor and taste. (8) The same oil as number 7, heated at 100° for one and one-half hours and filtered hot. (9) The same as number 7, heated at 100° for one and one-half hours, while at the same time a current of air was passed through the oil under partial vacuum. Filtered hot and bottled. (10) Fresh coconut meat, ground and pressed in a hand press to remove most of the milk. Afterwards this meat was dried completely by spreading it in the sun for about five hours. The oil expressed from this copra was almost water white and without taste and odor. (11) Coconuts split in halves and dried in the sun for five days. Ground and expressed. Yielded a cloudy, slight colored oil, very hard to filter, with a peculiar, but not unpleasant, taste and odor. This sample was strained through cloth but not filtered. (12) Same as No. 11, strained and filtered slowly through paper. (13) Same as number 11, heated at 100° for two hours and filtered through paper. (14) Fresh nuts, split in, halves and allowed to stand during one week in the air at room temperature (about 30°). A vigorous mold growth and an unpleasant odor developed. This moldy meat was dried in a vacuum and the oil was expressed. This was highly colored and was rather unpleasant to taste and smell. (15) Commercial coconut oil.shaken with 2 per cent of solid calcium oxide (burned lime), heated to 100° and filtered. The filtrate was treated with animal charcoal and again filtered; there resulted a colorless oil which was very free from an unpleasant odor or taste. (16) The same copra as that used for number 1; was allowed to stand one month longer in an open jar, then expressed. (17) Oil expressed from yacuum-dried copra which had stood for one month exposed to the air; the oil was heated to 100° and filtered. lS WALKER. (18) Expressed from sun-dried copra and treated in the same manner as number 17. Both of these samples were of as pleasant a taste as oils from fresh copra, (19) Vacuum-dried copra which had stood in a-closed desiccator over water for one month, and which had accumulated a very decided growth of mold. It was dried for one hour and expressed. The oil had a considerable color and was slightly unpleasant as to taste and odor. Heated to 100° and filtered. (20) Sun-dried copra treated in the same way as number 19. Yielded an oil somewhat darker in color but otherwise much the same as number 19. Filtered without heat. (21) Same as number 20, heated to 100° before filtering. (22) The same copra as that used for samples 1 and 16 was allowed to stand for three weeks over water and for one week in air, and then dried and pres A vigorous mold growth appeared in the copra and a peculiar ethereal odor was apparent. The oil itself was of a light-yellow color, with a pungent, rather unpleasant, odor and an extremely disagreeable taste. (23) Expressed from commercial copra, first quality, sun dried, Tacloban, Leyte. The unfiltered oil is dark colored and cloudy, depositing a black sediment. (24) Same as number 23, filtered. Almost colorless. ed. (25) Expressed from commercial copra, grill dried, Laguna (second quality) . Not filtered. (26) Same as number 25, filtered. Light yellow in color. (27) Expressed from commercial copra, grill dried, Romblon (considered second quality). The filtered oil is light yellow color. (28) Expressed from commercial copra, first quality, sun dried, Hoilo. The filtered oil is light yellow in color. (29) “Langis” coconut oil, prepared by the customary native process of grating the fresh meat, exhausting it repeatedly with water, and boiling down the emulsion thus obtained until it is nearly dry. The oil is then poured off from the brown coagulum which sinks to the bottom of the vessel. A freshly prepared oil, isolated in this manner, is very light in color and it possesses a decidedly pleasant coconut odor and taste. Before filtration it is more or less turbid, owing to the presence of a small amount of water and of albuminoids. (30) Same as number 29, filtered. The oil is water white. (31) Best grade commercial coconut oil, probably made from fresh meat. It is light colored, but very turbid and contains considerable water and suspendéd matter. : (32) Commercial coconut oil, probably made from copra. Very clear but highly colored. (33) Commercial coconut oil, Manila. Probably made from fresh meat. It contained considerable suspended matter and water. (84) Commercial coconut oil, Cebu. A highly colored siderable sediment in the bottom of the bottle. (35) Commercial coconut oil, Tayabas. A highly colored rancid oil made from copra. It is only a few months old. raneid” oil. Con- “ NOTES ON COCONUT, COPRA, AND COCONUT OIL. ood TasLe LV.—Percentage free fatty acid (as oleic). | Re |At start. Hane eee sates ae aes A 0.06 | 0. 06 0.09 0.60 2.6 9.6 2.3 8.9 B 0.06 | 0.06 0.08) 20.48 1 cera are |) at 1,8: |) — 8 1.5 1.9 3.1 9.4 2 1,2) 1,8 | 1.5 1.7 Dail 3:7 3 4 56 21 2.6 £3) Dasa 4 Sei |emeul ss I BO 6.1] 6.8 9.0 5 GSS fee |isralsteeneeri 7.6 10.1 | 28.2 6 0.10) 0.16 0.19 0.30 0.68) 1.9 7 0.16 0.18 0.19 0.27 0.39] 1.0 8 0:16) 0.14 0.19 0.30 0.40} 0.93 OO O96 0.18 0.25 0.35 0.93 10 0.16) 0.16 0.21 0.28 0.54) 1.5 ul 0,18)| 0.18 0.25 0.28 0.43 | 3.5 12 0,08) 0,00 0.10 0.14 0.28/ 1.0 13 0.18 0.09 0.09 0.15 0:28) 1a 14 3.5 Be ie GO ts che} 60 || ae} 15 0,62 [oo | ONSBu | Paes 353i | ence 16 ‘In LGR pest oe) lessen righifal Piaia Sei 17 0.09} 0.09] 0.14 0.16 0.25) 0.81 18 0:16) 018/025) 0.27 0.44] 0.82 | I 1.18 lu 1.34 1.58 2) 3.5 | 20 0.69) 0.69 0.74 0.85 Ld 27 | 21 0.69) 0.69) 0.74 0.82 isi 2.0 DOM Samo 343 fl lta abe Ne ersten |IRenale RW) [eee F983 14 1.6 1.8 2.0 26 | BO | 24 L4 1.5 1.7 1.8 25 | 6.0 25 2.6 3.4 3.6 3.9 AD Seal eee eee a5 28 | 9 26 3.1 3.5 dlap i Bee eran 27 oral 2.4 2.5 2.8 gil | Ao | 28 a0 | 8 4.0 47 6.1 | a. | | 29 0.08 0.38 0.60) 0.69 1.4 3.9 | | Bol - GO — GiB|) OnG) | O.10|) OspL | Bi 2.0 2.9 0g |e ae | 32 6.8 7.5 | 8 9.2°| 16.1 |p 33 5.5 8.2 | 16.5 34 8.7 Tie eG 35 55 Oke Mele 545 py] eeeeeererata eer er rexSi OE) has nha | | | | aThe greater part of the titrations at this period were made by Mr. L. A. Salinger of this Bureau who kindly continued the series during my absence from the Bureau. ) This oil was kept in a large bottle. A sample in a small bottle showed an acidity of only 0.09 a this time. c Large bottle. 4Small bottle. The most apparent fact which is noted on examining these tables is the proportionately large increase in free acid after the first year. In some samples this change is so great and so unexpected as to appear at first glance inexplicable, but in such cases I have nearly always 71978——3 : 134 WALKER. been able to detect some abnormality in the conditions under which the sample had been kept. For instance, numbers 1 and 2, samples of identically the same oil, the only difference between which lay in the fact that number 2 had been treated for three hours with a current of dry air to deodorize it, increased in free acid in practically the same ratio during the first six months; at the end of a year number 1 contains only 1 per cent more acid than number 2, but after three years shows 9.4 per cent as against 3.7 per cent in number 2. On examining the bottles in which these samples were kept I noticed that number 1, of which a considerable quantity had been prepared at the start, was still im the original 250 cubic centimeter bottle in which it had first been placed. About half of this oil had afterwards been taken out for the preparation of number 2, and filled into a 100 cubic centimeter bottle. This dif- ference in the size of bottle, then, with a corresponding different surface exposed to oxidation, must account for the 5.7 per cent excess of free acid of number 1 over number 2. A still more striking example of the influence or surface oxidation is afforded by sample B. This was an exceptionally pure oil, sterile and freed as far as possible from impurities by repeated filtration. One por; tion of about 25 cubic centimeters was transferred to a small bottle nearly, but not quite, filled, shortly after preparation, while in the original 500 cubic centimeter bottle there remained at the end of three years about 20 cubic centimeters of oil. The latter sample has increased in acidity 8.8 per cent in three years, while the former, from which air was nearly completely excluded has gained only 0.31 per cent free acid in the same time. Sample number 5, a commercial oil which changed from 5.5 per cent free acid to 28.2 per cent, had originally been treated with live steam to remove its unpleasant odor, and decanted into a bottle without filtering out all the water of condensation from the steam. When the final titration was made there remained between 5 and 10 cubic centi- meters of oil, together with considerable water, so that hydrolysis by water undoubtedly had much to do with the large amount of free acid developed. It will be noted that oxidation once started proceeds more rapidly in oil already having a large free acid content than it does in those com- paratively low in acidity (compare numbers 6 to 13 and 17 to 21 with the commercial oils from number 23 on). Exceptions to this rule are, as above stated, due to abnormal conditions of storage. The behavior of these oils during the period prior to the appearance of oxidation has been discussed in a previous paper and no new data have been brought out by longer storage which do not tend to confirm the conclusions drawn at that time. NOTES ON COCONUT, COPRA, AND COCONUT OIL. 135 CONCLUSIONS. The deterioration of a freshly prepared commercial coconut oil is produced by at least three entirely independent processes and may be divided into two distinct periods of time. The first, rapid splitting up of the fat, beginning immediately after its expression from copra and continuing for several months up to a year asioned by molds or more according to the nutrive matter present, is oc which are either pressed out with the oil together with sufficient sugars and albuminoids for their growth, or, in the case of hot pressed oils, enter the freshly prepared oil from the air. This action continues as long as sufficient nutritive material for mold growth remains in the oil. It may be completely checked by filtration, preferable after heating to 100° C. more thoroughly to coagulate albuminoids and to destroy any enzymes already secreted by the molds. Toward the end of this first period, oxidation by the air sets in and may continue indefinitely. The rate of this process depends upon the amount of surface exposed to the air, compared with the total volume of oil, and may in extreme cases cause an exceedingly rapid deterioration. It may be entirely prevented by storing the oil in completely filled recep- tacles, impervious to air. Along with the two above-mentioned processes, a slight hydrolysis due to heat, moisture and free acids already present is constantly taking place. It may be reduced considerably by filtration, which removes most of the water, together with the organic impurities. There is reason to believe that some hydrolysis is brought about by enzymes produced by the molds, as unheated oils which have been filtered and rendered antiseptic increase in acidity somewhat more rapidly than do heated ones under the same conditions. However, this distinction is not so apparent after the first year. Light has apparently no effect on the oxidation by air of coconut oil. PORTLAND CEMENT TESTING. By W. C. Rerpirne and L. A. Sarinerr. (from the Chemical Laboratory, Bureau of Science, Manila, P. I.) INTRODUCTION. This paper is devoted to a discussion of modern cement specifications and we have endeavored to point out many reasons why they do not exclude the personal error that is experienced by all testers working under them. Simple methods and precautions necessary to reduce this varia- tion have been suggested. ‘Throughout the discussion, the effects of the requirements of cement specifications, the difference possible in manipula- tion and the consequent variations in the results obtained have been il- lustrated. Certain powerful climatic influences that tropical conditions may exert upon cement are also discussed, and the last chapter suggests the characteristics which a cement should have to give the-greatest efficiency under tropical influences. SPECIFICATIONS. The “value of standard specifications to the engineer, the consumer, and the country at large is as great as their value to the manufacturer. A standard specification, provided it is both equitable and safe, cheapens the product, insures quicker deliveries and acts as a powerful regulator to the industries affected. * * * The danger of a fixed standard of any kind lies in tts becoming unprogressive and following behind the demands of the time.”* The last sentence should be especially emphasized. All official cement testing in the Philippines at present is done under the United States Army Engineer Specifications of 1902. No change has been made in these specifications in the last six years despite the great amount of work which has been done upon the physical and chemical properties of Portland cement in recent times, and, in the light of experience, it has been found that these specifications could certainly be improved. This unprogressive tendency is, perhaps, due to the inertia inherent in all committee work. Hach individual member has fixed ideas on certain questions, or on the results of certain personal experiences. As one writer 1 Editorial: Hing. News (1904), 51, 612. (Italics supplied.) 137 138 REIBLING AND SALINGER. puts it: “We are all disposed to argue somewhat on the basis of our prejudices or to refute others because of the prejudices which we associate with them. * * * Therefore it is difficult for us to arrive at conclu- sions purely by the Lght of reason, and to deal with every syllogism from its premises to its conclusions.” However, the American Society of Civil Engineers and the American Society for Testing Materials are constantly working to improve their cement specifications and these specifications will soon be adopted by the Government of the Philippine Islands for all civil and municipal work. They have accomplished much towards establishing a more prac- tical, impartial and comprehensive system of testing. Yet “notwith- standing that so much has been done towards unification of methods, it may never be possible to determine accurately the value of one cement as compared with another tested in a different laboratory”.? “Experience since the report of the committee was made has shown that the difficulties in the way of uniformity in such tests are much greater then was then imagined. The variation in the results of tensile strength between the work of different experienced operators working by the same method and upon the same material are frequently very large and often make all the difference between rejected and accepted cement. Differences in tensile strength with neat cement of 40 to 60 per cent are not uncommon, while for sand mortar they are much greater.” * At present all standard specifications leave much to be desired. A Government committee appointed to investigate the quality of a certain brand of cement, after much consultation with engineers, chemists, contractors and manufucturers, introduced its final report with these remarks : “There are no standard specifications which are regarded as absolutely correct. All tests are approximations and must be interpreted in accordance with the specifications in use, and with due regard to the purpose for which the cement will be used. “There is no practical difference between the qualities and properties of a rejected and of an accepted cement in the immediate vicinity of the limits set by specifications.” * It follows that the engineer may be in much doubt as to whether to reject or accept a cement. “It must be recognized, however, that cement specifications are not for average results, but are intended to cover the lowest limit which can be allowed in the work and to provide for lack of uniformity in testing as well as in real quality.” * * Sabin, Louis Carlton: Cement and Concrete. New York (1905), 30. * Spalding, Frederick C.: Hydraulic Cement. New York (1904), 115. ‘Final report of cement investigation committee appointed by Executive Order No. 60, 1907—The Government of the Philippine Islands. *Taylor and Thompson: Concrete. Plain and Reinforced. New York (1907). 99° PORTLAND CEMENT TESTING. 13% Such a conclusion may be satisfactory to the engineer, but should the tests be close to the margin specified for acceptance, the selling agent is sure to protest and order a retest of the material. He may allow the original tester to retest the cement, or he may send the samples to one of the many commercial laboratories whose reputation for high results in cement testing is well established. The retesting may produce satisfac- tory results in either case owing to the weakness of all cement specifica- tions. “It is not to be inferred howeyer that the highest results are neces- sarily the outcome of the greatest skill. As a rule the most expert and reliable operators get only moderate strength for the best material.” ° Such a condition of cement testing is very deplorable. Unless speci- fications guarantee an accuracy within 10 per cent, the greatest efficiency of a cement laboratory is also lost, as the mere mechanical routine testing of various brands of cement should be the least important part of its work and satisfactorily to accomplish the more important object, namely a systematic study of the peculiar effects of climatic conditions upon them, a variation factor of not more than 10 per cent is essential. Sabin states that “the chief object of testing cement is to arrange the various products in their true order of merit. Cement is at present used in a very crude way and it is only in exceptional cases that poor quality of material may be detected in the completed structure. This is suf- ficient reason why so few failures can be found in cement work which may be attributed to the poor quality of the cement. But in the more economical manner in which the material is, even now, being used, it is absolutely essential to know what its future behavior will be.” * We believe that the inefficiency of all American specifications lies in the fact that they do not outline sufficiently in detail the minor con- siderations and operations, and that to these minor details, owing to the peculiar and sensitive character of cement, is readily attributed a possible variation in the results of testing of 30 to 40 per cent. There are certain qualities in cement manipulation that can not be controlled, such as the size, shape and intermingling of crystals, nonhomogeneous voids in sand briquettes, unequal action of the water upon the hardening of briquettes, etc., but we believe that by far the greater variation is caused by the different manner in which different laboratories interpret the minor details of manipulation and treatment; and we also believe that if specifications were more explicit in this respect it would be sate to predict that different laboratories would agree within 10 per cent. This assumption is supported by the well-known fact that the system of the individual laboratory usually produces fairly uniform results, but a comparison between different laboratories which differ only in those details not explicitly treated in the specifications, often shows the most ° Spalding, Frederick C.: Hydraulic Cement. New York (1904), 160. 7Cement and Concrete. New York (1905), 82. 140 REIBLING AND SALINGER. startling inconsistencies. To be more explicit, we will give, for example, the following tests made under the Army Specifications for 1902: Two operators of considerable experience were ordered to test according to these specifications a shipment of 1,000 barrels of cement under dispute. itty samples, each representing one barrel, were taken at random, tested, and the figures tabulated as shown by the accompanying diagrams numbers 1 and 2 and by Table I. Tarte 1—Showing the variations from the mean of 200 breaks of each of the four sets of briquettes made by the testers. | | | | | gr cae | Neat (200 each). | REE Suey 3 (200 | e Tester. : aus = il | | | | 7-day. | 28-day. | 7-day. | 28-day | | } {eA SENS BL ERIE Ee) 502. 4 | 601.8) 143 | 220.6 | |) Sera es 566.8 | 641.3) 164 | 28.3 [Een sseenenees 534.6 | 621.6) 153.5) 228.0 | Difference_______ 32.2 | 19.8 | 10.5 | 7.4 ePeracent=seeeee +6.0 +3.2 | 6.8 | 3.2 | | Increase from 7 to 28 day tests. Neat.2 | Mortar, 1 to 3. v} | | | A. 3 | A. B. | ee | | FROWN CS peers | 99.4) 74.4) 77 71.3 Rerscents=se=se re eenes | 19.8 | 13.1 | 54.3] 43.5 aTnerease desired by specifications; 20 per cent. > Increase desired by specifications; 57 per cent. The fineness (through 100-mesh sieve) varied from 94.2 to 97.3. The specific gravity dried at 110° in all cases, was below 3.08 and ranged from 3.02 to 3.07. , Space will not permit of discussion in detail of the methods used by each tester, except to say that A used the automatic tamper described below and applied the blows differently than did B, who used the ordinary tamper; A’s briquettes were placed in a wet closet, B used the damp cloth; A’s briquettes were also always under water, kept running for a few hours each day; B siphoned the water from the pans and then refilled them, thus exposing the briquettes to air for about 10 minutes each time. All these differences were in accordance with specifications, as the minor details of testing were indefinite enough to allow them. One operator, A, always obtained lower results on an average than B, but B did not show the gain in strength between the 7 and 28 day tests that the cement was capable of. Evidently, B’s system gave abnor- mally high 7-day tests, and therefore 28-day breaks showed little increase. However, notwithstanding these differences, the variation between the samples themselves is also clearly evident. “LT (ON Nvaovid ONILLAS JO AWIL OS 64 8% Ly 94 Sy 47 Ew e7 Ib OF GE BE LE GE SE VE EC ZE IE OF. 62 BF Le 92 St ¥e EC 22 12 Of Gr Bi Lt 9 SI GI El Zi tO 6 8 £9 GF ¥ F&F ZT | Je | Tavs Jp WaISHAN | | | IES : —h 4 — Sie = + a | a] 7 ~ N a SS oo a \ Ww a m 5 < 4 HH — S 7 Sh at: ~ 5 | \ aa Ok NN = ~ se ee re 1 ‘ A yy \ Taint ry [ia eat /\_ hh ke iN 4 aL : vATS H \ a mee i\\ \ NATAL S ti \| \ ALAM [yn NV BD) | A=0 N04 \ i v ! \ Ny / “id TS ay" \ : t \ WMS ooz ~~ - PAINS SP - \ = Ts 1 ’ 2 \ Z 7 WATS ia] <6--0_ [7 y i VI] Adx =e ] 2) SS SS NLL Nie f f iH Se) ae Ai aa roa RS PT = s é \ | ome ? , : 5 IN 4 Soe ol : \ A} o£ ese i ai Gilwsalaalee = ar i | TVNily 3 ml a i] = [ tia4-— S PVN’ sy 00% "€ ‘ON ‘III “IOA “IOS ‘Nunor “I1Hg] : [DNILSAL INGNAD GNVILYOG :UTONITVS ANV PNITEITY REIBLING AND SALINGER: POR (PHIL. JOURN. SclI., Vou. III, No. 3. mn a III, No. 3. [PHIL. JOURN. SCI., VOL. ENT TESTING. ] REIBLING AND SALINGER: PORTLAND CEM fu YY TENSILE STRENGTH Dracram No. 2. \ PORTLAND CEMENT TESTING. 141 The diagram calls attention to the lack of uniformity in the results attained in each sample. The differences in the tensile strength are not only between the two testers, but also brought out by the same individual: those in the increase of strength with age, the failure on the part of one tester to obtain good breaks in many instances, and the great variation in the time of setting, are all apparent; yet both operators worked carefully and in strict accordance with the United States Army specifications. The committee in charge, after thoroughly investigating the methods of the two operators, reported as follows: “The methods followed by each cement tester differ slightly in certain details, but the differences are not regarded by the committee as important or as in violation of the purport of the specifications; and the results obtained in each laboratory are regarded as fair, equitable and reasonable to both manufacturer and user.” § : Even careful inspection of these charts would probably convince almost anyone that the tests, as they stand, are practically useless; that one of these testers, or even both, were inefficient; or that the cement itself was of a most peculiar quality. However, as will be shown, the whole inconsistency was due to a cause the elimination of which the specifica- tions do not even mention. It appears that the work of the testers for time of setting, 28-day neat and sand strength, and specific gravity determinations, was accurate and true to the quality of the cement at the time it was tested. The cause of the great variability so evident in diagrams numbers 1 and 2, was due to the fact that each tester worked the cement after it had undergone various degrees of exposure to aération, and that the influence of this factor produced very marked changes in the quality of the cement. THE EFFECTS OF AERATION. Most of the cement specifications now in use devote considerable space to the manner in which the samples should be taken, but they all neglect to state how these samples shall be stored and preserved until tested. As a result, they may be sent to the laboratory in wooden or paper boxes, paper or cloth bags, tin cans, galvanized-iron cans, glass jars, etc. The cloth and paper may be thin or thick, and the cans, jars and boxes may have tightly or loosely fitting covers, or even no covers at all. These samples may be tested as soon as they are received at the laboratory or, owing to the amount of routine work already on hand, they may stand for some days before being worked. As a result of all these conditions the cement may have been. subjected to unequal aération and its charac- teristics changed accordingly; this change has often been sufficient to alter the resulting tests from satisfactory to unsatisfactory. 8 Final report of cement investigation committee appointed by Executive Order No. 60, 1907.—The Government of the Philippine Islands. 142 REIBLING AND SALINGER. The literature on the subject of cement is filled with information on the effects of aération.® which are “unsound” due to the presence of free lime, are improved by exposure to the air, but cements high in Cements high in lime or those alumina, especially if lightly burned, are apt to become. quick setting and otherwise dangerous under the same treatment. This is especially true in the tropics as “aluminous cements are readily subject to alteration in surroundings exposed to alternate dryness and humidity and also when exposed to a high temperature.” 7° Cements are encountered, the fineness and soundness of which may be very satisfactory throughout, but the specific gravity, time of setting and tensile (thes tors: to the None of the pats may warp or disintegrate, even during steam and air strength mortar especially) may vary from one extreme other. exposures, so that perfect soundness may be a characteristic of such material. In fact, it is possible for a pat to remain at a red heat for several hours before it disintegrates in any marked decree. Cement of this class, according to chemical analysis made from time to time, proved itself to have a uniform composition in The It is hardly necessary to add that every known precaution was taken to secure uniform all respects except the loss on ignition, which varied from 1 to 6 per cent. silica content was uniformly low and the alumina and iron high. results. Experience in this laboratory has demonstrated that in most instances variations such as those mentioned are encountered in cement samples which have been received in thin paper bags, or which had otherwise been exposed to the air; and that cement received for testing in closely covered cans and boxes and not subsequently exposed to the air, usually gave very acceptable, uniformly good results. These conclusions are emphasized by the following tables: TapLe I1.—Characteristic examples of tests of cement stored in cans.” Mortar, 1 to 3. | | Mortar, 1 to 3. Sample No. = RE SES SF Sample No. | 7-day. | 28-day. | 7-day. | 28-day. | - ——| 240 | 289 | 995 | 307 210 | 238 295 | | 210 | 205 | 299 | | 193 | | 212 260 | 210) 315 | 248 | 288 | | | | ® Meade: Chem. Eng. (1907), 5, 341; Taylor and Thompson: Conerete, Plain and Reinforced, New York (1907), 62; Spalding, Frederick C.: Hydraulie Cement, New York (1904), 4, 56, 80; Candlot, M.: Cement et Chaux Hydrauliques, Paris, 1891. Spalding, Frederick C.: New York (1904), 81. “The setting time in all of these cements was satisfactory and uniform. Hydraulic Cement. PORTLAND CEMENT TESTING. ~ 14833 Taste III.—Variations in the same brand of cement received in paper bags. {Mortar, 1 to 3.] Highest. | Average. Lowest. | Sample Sample | Sample No. i| No. | No. | 7-day. |28-day.) 7-day. | 28-day. | 7-day. | 28-day. | | i | | DI-1 187 | 269 || CE 168 260 || B91 126} 194 | DI2/ 179} 141} 2141 ||" B9-2 166 | 240 | D1-3 | 191 | | 141) 210 || B93 140 | 181 ‘| Dia] 180 | | 189} 232) “Boa 143] a7 | D5 | 169! 234 5-5 | 144 | 202 |) B9-5 | 106! 125 DI-6 | 176) 245\|| 5:6 195 257 || B9-6 138 | 193 | =f eel 6S ene? | en CE 180] 238'|| B9-7 135 | 186 D8] 184] 257 || 5-8 145 924 || B9-8 175 | 178 D9) 174 234 || ©5-9 168 204 || B9-9 103 | 181 D1-10} 182} 234 || ©5-10 145 210 | B9-10 148 | 185 | | | HI TavLe 1V.—Time of setting. - Initial | Final set. set. Initial | Final Roth Ret Sample No. Sample No. 3 ot o SO oo im Ss i=) Co. oo Ss oo Co oS The above results led us to investigate more specifically the effect of air exposure on this class of cement. In Table V the depreciating effect upon the mortar of freely exposing a small amount of cement (about 400 grams) to the air in open jars for ten days is clearly shown. All mortar mixtures made from cement fresh from the sample package, dried or undried, and exposed to the air not longer than eighteen hours, did not set before the molding was completed and passed in tensile strength. All mortar mixtures made from cement exposed to’the air for ten days, set before the molding was completed and therefore failed in tensile strength. 144 REIBLING AND SALINGER. TasLE V.—Test with dried and undried cement taken from paper bags. [Mortar, 1 to 3. Temperature of room during molding and setting, 29° to 30° C.] ] | Sample | A P | Pounds | No. | Sasa | Condition of mortar = per Water (same | Condition. | when molded. | Ase. | square | used. brand). | | | inch. | | | | | | Days. Per cent. F3-9 | Dried ana exposed to ait 10 days____| Set in 10 minutes_____! a 79 | 12.5 F3-9 | Dried and exposed to air 10 days___-) Set in 12 minutes_____ 7 | 98 14 F3-9 | Dried and taken fresh from sample | Not set in 20 minutes. | 7 | 186 | 14 package. | F4-4 | Dried and exposed to air 10 days___-| Set in 12 minutes_____ | 7 | 70 | 12.5 F4-4 | Undried, fresh from sample package_) Not set in 20 minutes=) 7 182 | 12.5 F4-4 | Fresh from sample heated at 100° | Not set in 20 minutes_, 6 157 | 12.5 C., exposed 1 day. | | | F4-4 | Fresh from package undried_______- Not set in 20 SEY DETES u | (@) | 12.5 | F4-4 | Dried and exposed to air 10 days__._| Set in 10 minutes_____ | 7 92 12.5 F4-8 | Dried and exposed to air 10 days____| Set in 12 minutes_____| 7 | 101, 14 | | | FA-9 d | WG | eee ‘Sunariea, fresh from sample package_} Not set in 20 minutes_! ” = F4-10 | | 178 | 12.5 j | | 88 briquettes—146 lowest 183 highest. Percentage of water evaporated from F4-4 in drying was 0.66. In all cases not specified the highest of four good breaks is recorded. These results demonstrate that heating did not cause quick setting, and low tensile strength, but exposure to air in open jars for ten days did. Table Va illustrates the same action: TABLE Va. [28 days; mortar, 1 to 3.} | | Sample No. | Water. Stored. see piver.8 pTeuay | ; | Per ct. | 123 | Covered can_____- 315 | 302 | 295 | 12: | Uncovered can __- 247 205 | 158 | | 121 | Covered can.__-.- 300 2298 202 | 12: | Uncovered can __- 245 198 7 I | «Cover removed from can on thirtieth day. PORTLAND CEMENT TESTING. 145 The samples included in the next table were received at the laboratory in paper sacks on the 27th of the month. The next morning the two sacks comprising each sample were screened and well mixed. One half of each sample was then put into a tightly covered, galvanized-iron can, and the other half put back into the original bag. The time of setting was taken after seven, eleven, and twenty-four days with the results given in Table VI. TABLE VI.—Comparison of results of samples in paper bags and in closed cans.* | | i Worked after 7 days. | aaa ‘Time I gauged. | | TR ie Sample Sioned | Water. ee eater of Gauging.| Setting. | |—— | | | A.M. | Per ct.| h. m. °¢. | 8.10 1|Can__| 21} 1 10| Plastic__..-____ 27 «| 27 -29.5 8.20 2 | Bag___ 21 0) |e Se 7 | OF 5 | 8.30 3 | Can___ a] 1 27 | 27 29.5 | 8.40 4|Bag| 21 27.5 | 27,5-29.5 8.50 5 | Can___ 21 27.5 | 27.5-29.5 9.00 6 | Bag___ 21 27,5 | 27.5-29.5 9.10 7|\ Cam) | Bl || Le TO J LO eee eae | 781598 27915229115 | 9.20 8|Bag| 21 M0) |e Gee ewes | 98 | 98 =2915 9.30 9] Can) 21 Bone Gem cD | 98. | 98 -29.5 9.40 10 | Bag___ 21 15} |e domes 98 | 28 -29.5 Sample No. 8, worked after 11 days. pose aut | Bi Gaa 4) Oil! wm a0 27.5 | 27.5-30 | 8|Bag--| 21 17 | 27.5 | 27. | 8) (Camas 225) oler25 a 27.5 | 27.5-30 8 | Bag___ 22 20 27.5 | 27.5-30 | 8 | Can___} 24.) 1 30 |)Poo plastic toy 27.5 | 27.5-30 | 8| Bag__| 24 | 38 } hold its shape. | 27.5 | 27.5-80 i 1 Worked after 24 days. | | 8.80 2|Bag__| 22 91 | Very plastic-..--.| 28 | 28 -30 8.40 3 | Can___ || @ B's lopment 28° | 28 -30 8.55 3 | Bag___ OB OO Nes 0) ape | SAY) 9.05 | 4) eyes) op) at ep | 28 | 28 -30 Q6 | 4 Cam] 2a!) @ 80 |e | 23. | a8 0 9.25 | 8 Can__- OD ent30)\|eeaeee domme. le er25iealkestr=30 9.35 | 8 Bae | 22) | ear 9) |e a ee om | 98 | 98 -30 «All pats were made by the same operator by the Gillmore needle method described below. Before weighing, the original samples were thoroughly mixed. 146 REIBLING AND SALINGER. Chemical analysis of number 8 sampled on the eleventh day gave: TanLe VIl.—dAnalysis on-eleventh day. | | From From Constituent. | bag can | Per cent. Per cent. 19. 80 20. 24 8.3 8.50 2.75 | 2.98 | 63.44 | 63.32 | Magnesium oxide ___ 2.25 2.15 [ee Moisture; (1! 0S) See ee eee 0.32 | 0.14 WSS (Oh TYSON MOO ee fo a Ba — Sbp Sulphuric acid (SO;) ___ 0.43) 0.42 Taste VIII—Woistwre, loss on ignition and carbonic acid. | On twenty-fourth | | day cement from—_ Sample No. 8. | Can. Bag. Per cent. | Per cent. | Per cent. | Moisture at 110° 0.14 | 0.40 | 0. 26 | Loss onignition after drying__ 2.63 | 3.92 | 1.29 | Carbonic acid (CO,)__-__--------__- 1.14 | 1.64] 0.50 | Combined water ___--__---__----__- Pe 3620) S| = ha The rapidity with which this absorption of carbon dioxide and water may take place (the local climatic conditions being those of the early dry season) is shown in Table IX. Two different brands were investigated, and in each case 50 grams of cement were taken from each of the samples specified, and put into 100 cubie centimeter beakers. These were accurately weighed and the free moisture then thoroughly driven off by four hours’ heating at 130° C. The beakers were then allowed to stand in the balance room, open to the air, but protected from dust by paper coverings. The gain in weight was noted at the intervals of time designated. At the end of twenty-eight and one-half days, the moisture was again driven off by continued heating at 130° C. and the amount of water absorbed subtracted from the total absorption. The samples were again reheated after thirty-five days’ additional exposure. The results obtained in detail are as follows: 147 a te AND. CEMENT TESTING PORTL “UB UL PaAtadayy q *SBq UL paatadoyy v 68°6 690 TOT 16 'T €8 0 [§'0 | GF 0 | L460 | ILO | ILO | 0L°0 | GL°0 | Te°0 | 11:0 | 21°0 | FL‘0 | g0°0 | 80-0 OTC OM SARS ON e| pase aie noe aaa ar aLI@T | % SPS 19°0. eL'T 00°% TsO 48°0 | 6F'0 | 8F'0 | ILO | ILO | TL‘0 | ZL'0 | 620 | FLO | ZE‘0 | Gt‘o | ¢o‘0 OO C1 | as Om | ca asIeT | & OF @ 6¢ 0 GOT L461. 80 6&0 | 6F0 | 9F'0 | ITO | ILO | OL°0 | ZL°0 | 820 | F1‘0 | zT'0 | gt-o CNW) SOW. || CHEW GIB) | PRS Ss sesereecs a80cT | Z POG LL°0 S$a'L 91°T 960 0§'0 | &F'0 | &F0 | O10 | O10 | OL'0 | TL‘0 | 080 | 2L‘0 | zt‘0 | 910 | Foo STA LEROVE |AG Tees O28 115 Zit Ogee ii cae aan v6-94 | T 61% 080 oT FLT TO'T Tg 0 | €F'0 | OF 0 | OL‘0 | OL’0 SHOP BSG OF ROSNOM Zn OM M/sTy ON CTs Oe iGO MOM In Ollie ON (61/0 O Se | aan ina as vl | T CPG 89°0 qi TL 6 °T 00°T FE ‘0 | SFO | 4h°0 | ILO | TL/0 GON RST NOR ECSHONI NOT) MG ON STON sFOROMRGEMON RCT: ON | 0/2): O lalla ine rG-94 | T 6 °T 84°0 TEV COL OL T G80 | SF°0 | tF'0 | OL ‘0 | OT"0 GON EIOS FCG iOF ESTs ONO TON PLLON | CONO!| RGLEON|ESTS Ott OE OM | sama unin aun wna sama v§-90 | T 10% 8L4°0 LOVT Lod LOT G60 | FF'O | SFO | OL'0 | TL0 CRON ETON ROS O Nee ORIG TOn RCTs ON eZOKO8/ECTRON| NG [a Oks| eh pO Ms iain ease em inceeces wL9A | T Yd Vid | Pid Pd FOL | PDF Lh | 49 | 99 ed | 99 | 99 | 99 | 99 | 99 | 99d | 90 D | 19d | 2d | 0d | 99 ’ Q le! ape =e Z 3 5 Oi @ fe lac hae tose | ep Ge | se Pedy deeb a =8 z Bases | AES I aigp Be 5B 8 eof Re FP = S oe 7 3 oge | ag os 2 5 ao ae 3 z Re m a z ia © a Sit) 5 os he) Bo “SARL *sInoyy es $ Bo | os | Se | od | of i peed) aie | Se | Zen) Be a Al : SEs leeks on ordumg og oS {ag ayee ge iS | = ae || | eg re) oe BS rss om bs = a oe ae | wu as 2 3 “UORIpPAY JO 1B pus ag acs | eit 7 as 2 ‘OANASIOW OPIXOIP UOGIBI=ITR 07 aIMSOdxXd UO ISTO UT UTR eis Be Q gL iS) I a “UD 07 atnsodra uo py bian UL sabunyg— Ky Way 148 REIBLING AND SALINGER. TABLE X.—After exposure for sixty-five and one-half days. | Carbonic | Loss on acid Sample No. | ignition. | | | Per cent. | Per cent. Ian || The rapidity with which this action will progress depends upon climatic conditions, upon the nature of the cement itself and upon the ratio between the volume of the cement and the surface exposed. The action may take place very rapidly on the exposed surfaces, and yet penetrate into the mass very slowly. A well-mixed cement was stored in a uncovered can for one month. Cement taken not lower than half an inch from the upper surface set in twenty minutes, while cement taken 6 inches below this surface did not set until two hours. It is therefore absolutely essential in order to secure uniform results from the same sample thoroughly to mix the cement before weighing; otherwise a wide dis- crepaney between the specific gravity, time of setting, and tensile strength may result which could not otherwise be accounted for. The above data are considered suflicient to illustrate the effects of aération, although in this laboratory we have many more experiments proving the same facts. It may be well to state that cements have been encountered which do not change to any appreciable extent after exposure for several weeks. It has been shown (‘Table VI, can 8; eleven and twenty-four days) that cement, otherwise susceptible to a marked change by exposure to air, when preserved in closely covered galvanized-iron cans will be little affected by storage; and that thin paper bags do not eliminate the atmospheric influences. It has also been shown that the characteristics of a cement often undergo a change upon exposure which may be sufficient to make failures of otherwise acceptable tests. Therefore, it is evident that no system of cement testing, however accurate, will insure uniform or even comparative results until a proper, specified preservation of the samples after they are taken from stock is made compulsory. As a result of the considerations given above, it would seem necessary so to modify the ordinary procedure that the quantity of the cement deemed necessary for the desired tests should be freely exposed to the atmosphere of the laboratory for seven days in a layer 1 inch deep, in order to determine the effects of aération. A comparison of the-specific gravity, setting time, and loss on ignition of the cement before and after exposure, will give valuable indications as to its nature. The determina- tions made before exposure will be consistent with the quality of the cement at the time the stock was sampled, and the second treatment will show the qualities liable to be developed by subsequent storage. If the PORTLAND CEMENT TESTING. a 149 effect of exposure is considerable, troublesome variability of the brand may be readily attributed to this cause and the manufacturer will then know how to improve his product accordingly. Changes caused by the atmosphere penetrate very slowly into the mass of the cement in barrels, because the wooden staves and heads keep the air from the product to a considerable extent and the mass of material is large, therefore, alterations which may develop very rapidly in a small sample exposed in the laboratory would not take place in the barrel until a much longer period of time had elapsed. Spalding ™ states that “the effect upon cement of retaining it a long time before using depends upon the nature of the cement and the method of keeping it. When the cement is inclosed so as to prevent the access of air, as in barrels, it may usually be preserved for a considerable time without experiencing any alteration, provided it is kept dry.” The fact that cements stored in good barrels undergo very little change by a month’s seasoning is illustrated by the original and the re-test of the following cement, the results of which are recorded in Table XI. The samples for the re-test were taken a month after the original ones. All the samples were protected from aération before testing, and the re-test gave only slightly lower results, although sub- sequent experiments showed that the cement rapidly deteriorated in time of setting and tensile strength when subjected to air exposure. TaBLeE XI.*—Original and re-test (one month later) of cement stored in barrels. ORIGINAL TEST. Ie . Tensile strength, ‘ He alee Tensile strength, neat. seStnoxtans at, ine- nitia 1 Sample No. ness. Sp. gr. set. set. 1 day. 7 days. | 28 days. | 7 days. | 28 days. h. m.| h. m. oA ee ater 95. 8 3.07 2 10} 5 00 392 516 640 240 285 Dose. 96.5 3.07 2 On| Ome LO) 351 563 605 200 278 See 96.2 3.08) || 2 10)!' 5 00) 354 532 687 221) 300 Quran te Use 96.0 3.08} 2 10) 5 00 351 558 610 232 296, Obese 97.0 3.08) 2 00} 5 00 312 542 653 222 300 Average_| 96.3 eb.) BO) & B 352 542 | + 639 223) | 292 RE-TEST ONE MONTH LATER. 1 eee eel 94.5 3.08 | 2 25) 4 20 322 570 594 210 283 Detected 94.0 3.08} 2 20| 4 15 349 566 621 195 | 278 Sie ae 95.0 3.08 | 2 20] 4 40 312 531 583 200 | 300 Aes Re See 94.7 3.08 | 2.10] 4 35 325 559 612 | 199 | 281 Average — 94.5 3.08 | 2°19 | 4 28 327 556 602 | 201 285 aThe soundness of all samples was satisfactory. 2 Tbid., 67. 71978——4 150 ® REIBLING AND SALINGER. Many engineers maintain with good judgment that a cement should not develop dangerous properties on exposure to the air and if it does so it merits rejection, especially if the unexposed samples show irregular- ity, since it is only practicable to test one barrel out of every ten or twenty in stock. It is also true that barrels often become broken in shipment, and should the cement which is so received develop dangerous properties, the strength of a whole structure might be weakened by its use. Portions of the original samples of the cement the tests of which are recorded by diagrams numbers 1 and 2 were preserved and the tests after aération in paper bags are shown in Table XII. All of these tests were manipulated in the same laboratory, and they show that this cement is more or less readily subject to the influence of aération. Taste XII.—Test of cement shown in diagrams numbers 1 and 2, after aération in paper bags. vies : Specific gravity after— Initial set, in minutes, after ARGO GH storage in paper bags for aot oe | the times given. pet? ‘ ; Heating | Sample No. ture dur-| Storage in bags. a g | ing set- to red (nhayeg, «|e | en 28 F Andke z opie Pee eA weeks’ | 1 day | 34 days.*| 23 weeks.» 34 days. | 23 weeks. exposure.| EXC Ee Oo ess ak 220 140 45 PES) 3.02 3.14 EXD) Rat PRR 130 | 110 65 27-30 3. 06 3. 03 RG ers esr once 120 | 85 30 27-30 | 3. 02 3.02 3.14 | «21 per cent water. »22 per cent water. THE DISPOSAL OF CAKED CEMENT. It is usually specified that cement shall be screened through a 20-mesh sieve and thoroughly mixed before testing. The object of the sieving is to break up lumps and remove wood splinters, stones and other foreign substances. Such a procedure is of course proper. However, under certain conditions, the disposal of caked cement when it is present in considerable quantity should be more fully described. Cement literature has repeatedly pointed out that the tensile strength of a caked cement is considerably below that of the original material before it formed lumps. ‘This is due to the absorption of moisture and the subsequent partial setting. The difference that may result from such a change is illustrated in Table XIII, which gives the tensile strength developed by two samples of the same cement, one free from lumps and the other caked. PORTLAND CEMENT TESTING. 151 TABLE XIII.“ | Condition. * Mixture. | 7-day. |28-day.| 3 months. | l year. | 2 years. | 3 years. eT | | Alllumps.__—_____. IN Galt soeneest aeetcs (Be it 589 690 | 705 739 719 | | ENO pLur mn ps! | eS eee es 756 798 858 857 805 | ANT ilu reny oe WOE oe 1381 | 244 326 | 373 | 372 373 IN@) Ua Na 192 | 330 330 | 430 | 449 450 1 1 The disposal of these lumps then, especially those which are too hard to be broken up in the process of sieving, may exert considerable in- fluence in the tensile strength obtained. If one tester pulverizes the hard lumps and mixes this powder with the original sample, and the other simply throws them away, uniform results can not be expected. Should the cake be present in sufficient quantity to affect the tensile strength appreciably, the person requesting the tests should be notified of this condition. The presence of the cake may not be due to any fault or carelessness on the manufacturer’s part. lmproper storage while in the hands of the engineer or contractor may have caused it. Instances are also on record in this laboratory where caking was induced in the samples after they were taken from the stock. The samples were taken during a rain storm and through carelessness and imcompetency on the part of the one handling them, they were allowed to get wet. If specifications are to guarantee uniform and just results in all cases, the treatment of caked cement must be more fully described than it has heretofore. INFLUENCE OF TEMPERATURE ON TIME OF SETTING. The general rule for all cements is that increase of temperature in- creases the rate of setting. However, there is no fixed ratio between the temperature increase and the accelerated setting produced by it. The published reports of skilled operators vary in this respect, and L. Tetmaier,* after years of the most careful work was forced to adinit that “different cements are differently influenced by alteration of tem- perature * * * and it is scarcely possible to deduce a general law for eyen one class of cements.” The results we have obtained on the setting time of various cements, worked in the cold-storage room and at local temperature, have shown that the samples, in this respect, could be divided into three classes : 1. Slow setting cements, little affected by a variation of temperature from 20° to 30° C. 18 Griesenauer, Hng. News (1906), 55, 68. U Soc. Chem. Industry (18938), 12, 1036. 152 REIBLING AND SALINGER. 2. Cements which are slow setting at 20° C. but quick setting at 30° C. 3. Cements which are quick setting at 20° C. and also at 30° C. As these results were obtained while we were endeavoring to determine the variation which this climate might cause in tests which according to American standard specifications are at about 20° C., we will publish them in full in Tables XIV, XV, XVI, and XVII. In all the following determinations the cement was sieved and then very thoroughly mixed. To eliminate any elect of unequal atmospheric exposure each sample of 500 grams was put into a dry, clean bottle and tightly corked until used. In gauging, the water was allowed to soak in for one minute, and then the paste was vigorously troweled for four additional minutes. TABLE XIV.—Olass No. 1, slow-setting cements, little affected by variations of temperature. $ |Temperature when made,| Te™PeTa- A ture of | Initial| Final | Condition of 2 - room dur-) Water. | “cot, set. paste. = ing set- iS Cement. | Water. | Room. ting. °C. °C. °C. °C. Per cent.) h. m.| h. m. 1 19 19 19 19 -19.5 20.66 | 2 35| 6 15) Slightly plastic. 2 19 19 19 19 -19.5 21.66} 2 45] 6 15} Plastic. 3 19 19 19.5 | 19 -19.5 22.66 | 3 00) 6 20 Do. 4 19 19 19.5} 19 -19.5 23.66 | 3 00] 6 20)! Very plastic. 5 32 32 31 82 -32.5 20.66 | 2 35) 6 20! Plastic. 6 28 28 28.5 | 28.5-31 21.66) 2 20) 4 20 Do. 7 28 28 28.5 | 28.5-31 22.66 | 2 2 4 30 Do. 8 28.5 28.5 28.5 | 28,5-31 23.66 | 2 55 | 5 85] Very plastic. 9 29 29 30 30 -31 23.66) 2 25) 4 15 Do. TaBLE XV.—Class No. 2, slow-setting cements at 20°, quick-setting at 30° C. S |Temperature when made. Tempera- ate room dur-| water, |7aii#1| Eigal | Condition of s Cement. | Water. | Room. ting. °C. °C. °C. °C.” Per cent.| h. m.} he m. 1 31 31 31 81 -33 21 (a) (a) 2 31 31 31 31-33 22 25 | 2 30] Slightly plastic. | 3 28 28 28.5 | 28. 5-30 21. 66 21 2 50 Do. 4 28, 28 28.5 | 28.5-30 22, 66 28] 3 00] Plastic. 5 28.5 28.5 29 29 -30 23, 66 32} 3 10 | Very plastic. 6 19.5 19.5 19.5 | 19.5-20.5 21,66 | 2 32] 3 40} Slightly plastic. 7 19.5 19.5 19.5 | 19,5-20.5 22.66 | 3 380] 4 40} Plastic. 8 19.5 19.5 19.5) 19.5-20.5 23.66} 3 50} 5 00/ Very plastic. 2Tmpossible to make pat. To determine the effect of gauging in the cold-storage room and setting in the laboratory, and vice versa, two pats were made from each paste and one subjected to the change in temperature; the results are recorded in Table XVI. PORTLAND CEMENT TESTING. 153 TABLE XVI. Bot! pat Se eee aan Initial| Final | Condition ue. No. ie ae WENES sete set. of paste. Cement. | Water. | Room. ting. oC! Yas oC: CXor | Per cent.| h. m.| h. m. Gi 20.5} 20.5] 21.5} 20,5-22 22/66 | 2 15 |___-- Plastic. 1) 2 20.5} 20.5] 21.5] 32 -32 22. 66 55| 2 10 Do ee 20.5| 20.5) 21.5} 20.5-22 G31) DB AW | Do 2 t 2 20.5) 20.5] 21.5] 32 -31.5| 22.66 60 | 2 30 Do oa 32 20 32 | 32 -31.5| 22.66 26| 1 55 Do. 3 { 2 32 20 32 | 20,5-22 22, 66 BO |e Do 1 32 32 32 | 32 -31.5| 22.66 21) 1 21 Do é { 2 32 32 32 | 21.0-22.0| 22.66 38) | oem Do TABLE XVII.—Class III, cements which are quick setting at 20° and also at 30°.° Tempera- emDeres Bottle No. Cement. eee wag GE Water. Te Sorsaee of cement, ing set- i 5 and room. ting. oC. °C. Per cent. | h. m. A 17-18 22} 1 40) Plastic. A 29-30 22} 1 20 Do. A 17-18 21 n45) Do. A 29-30 pal |. al Pit) Do. A 17-18 20 40 | Slightly plastic. A 29-30 20 45 Plastic. A 17-18 19 | (a) (0) A 29-30 19 (8) (8) B 17-18 24 28 | Very plastic. B 29-30 | 29 -31 24 24 Do. aImpossible to make pat. Remark.—Pastes made with 20 per cent of water were more plastic when made at 30° than at 17°. k MOIST-ATR CLOSET. Another possible source of error which may be accountable for con- siderable variation may develop during the moist-air treatment. Most specifications allow the briquettes to be stored for the first twenty-four hours in a moist-air closet or under a damp cloth. Moist- air closets are given the preference, as unequal drying often occurs in using the damp cloth. A well-constructed moist-air closet is essential to uniform results. One condition in the use of a moist-air closet that is liable to have considerable influence upon the result of the cement tests should be taken This table also shows the marked effect that 1 per cent of water more or less, will produce upon the plasticity and time of setting of some cements. 154 REIBLING AND SALINGER. into consideration. This is produced by the heating of the cement after it is gauged and molded. Many cements heat considerably at some period of the stage of early setting and hardening, the rise in temperature often being as much as 10°C.; it may take place in five minutes, or it may not occur until many hours after the gauging. C. Prussing *° states that “many slow setting cements of excellent quality begin to set after five or six hours and then set completely in one hour, giving a rise of temperature of 5° to 7°C.” The heat gen- erated by briquettes placed under a damp cloth is not confined, as it is readily conducted away into the surrounding atmosphere. Moist-air closets are constructed to imsulate the interior from outside heat in- fluences as much as possible, and as a result the heat generated by the briquettes is confined; so that in a cubical moist-air closet of 2 feet on the side which was used to store briquettes after they were removed from the molds, the temperature often rose to 40° (30° being room tempera- ture) when it held from 80 to 100 briquettes. A number of slow- setting briquettes made at different, successive intervals of time, or worse still a mixture of quick, normal, and slow-setting cements, will under these conditions not be subjected to the same uniform temperature, or to a temperature change that is characteristic of it during its most critical setting and hardening period; and the same is true of the storage of pats made for the time of setting when many are placed in one com- partment after gauging. It is well known that the temperature conditions under which cement sets and hardens will influence its "tensile strength. Therefore, the practice of storing numerous briquettes in one compartment of a moist- air closet is very liable to cause abnormal one, seven and twenty-eight day breaks of some of them. Pats for time of setting similarly stored may be also affected to such a degree that an otherwise slow-setting cement may become quick setting. We suggest two ways of overcoming this objectional feature of the ordinary closet. The heat generated by the setting cements may be conducted away by means of a forced ventilation of air saturated with moisture; or only briquettes and setting pats of the same cement made at practically the same time should be placed in a small, msulated com- partment. The former method will maintain the interior of the compart- ment at nearly the same temperature as that of the laboratory, while the latter will meet the conditions of actual service, as the heat generated by the cement is not readily conducted away. Laboratory tests should coincide as closely as is possible with the actual conditions of construction work, where large volumes of concrete are tamped into wooden frames. The heat generated in such a large mass (especially in the center of it) is not conducted away by ventilation and it is in fact partly insulated 6 Thonind. Zeit. (1894), 18, 251. PORTLAND CEMENT TESTING. 155 by the wooden molds. Therefore, a moist-air closet formed of several small, insulated compartments, each with its own pan of water, is best adapted for the purpose. Such a moist-air closet has been constructed for this laboratory. Standard cement specifications should include a definite form of moist- air closet with a complete description of the materials for its construction, its dimensions, and directions for its use; otherwise one source of the so-called personal error will persist. TIME OF SETTING. The American Society has adopted the Vicat needle method for de- terminations of the time of setting. It seems to be the general impres- sion that the Gillmore method does not insure the desired accuracy, and many cement testers will regret that such a convenient and time-saving process has been supplanted by a more cumbersome one; still the Gillmore method, if properly regulated, can be made accurate, rehable, and im- partial and at the same time retain its simplicity, even though the meager directions in the United States Army specifications do not insure uniform results between different operators and at times imposes unjust tests upon some good cements. Merz, Meyer, Schiffner, Bohme and many others have each pointed out that to determine the time of setting of a cement it should be gauged with a quantity of water proper to it.‘ It has often been demonstrated in this laboratory that 20 per cent of water is not enough to meet the re- quirements of the fineness, specific gravity, chemical composition, and physical properties of many good Portland cements sufficiently to produce a paste plastic enough to be molded into a pat. The resulting paste is often so dry and non-cohesive that it will not stick together or to the glass plate; and yet 1 to 3 per cent of water in addition will produce the desired plasticity and cohesiveness. The whole phenomenon of the manufacture of artificial stone from finely powdered cement is one of solution, hydration and subsequent crystallization. The addition of sufficient water is essential for proper solution and hydration. The addition of too much water is to be avoided because of its effect upon the subsequent crystallization, and because the density of the paste must allow of proper manipulation. “Therefore, it is very evident that plasticity and not a given percentage of water should be the condition regulating the paste for cement pats. The insistence of the United States Army engineers upon a paste gauged with 20 per cent of water seems to be a striking illustration of Spalding’s assertion that ** “tests may be imposed which in nearly all “Soc. Chem. Industry (1891), 10, 928. 8 Tbid., 87. 156 REIBLING AND SALINGER. cases will secure good material, but often at the expense of rejecting equally good or better material.” Merz, Meyer, Schiffner, and others also insist that even when cement is gauged to the proper plasticity there is a large personal error due to the operator himself. After a careful study of this personal error, we have come to the conclusion that it is mainly due to the following five causes : 1. The manner of applying the needles. 2. The presence of small air bubbles near the surface of the pat. 3. The difference in the amount of water brought to the surface in patting the cement together and its presence there in a more or less liquid layer. 4. The difficulty in judging the exact time when the needles cease to make a “visible impression.” 5. The difference in plasticity. To overcome the first difficulty the pat should be made with a flat (not rounded as specified) top as illustrated in fig. 1. The needle should then be applied very gently and after the flat poimt rests upon the surface of the pat the full weight of the needle should gradually be applied. Failure to hold the needles in an exactly vertical position will often cause the edges to indent where the flat point would not. VELL LTE ESTEE LPB ELLEEDEEL ELDEST LLY PL MD (eS 9 ee Fig. 1. To overcome the second, third and fifth difficulties, the cement is gauged with the least amount of water which after one minute’s soaking and four additional minutes of vigorous troweling will produce a paste sufficiently stiff to retain its shape, and yet so plastic that the initial needle will sink almost to the glass plate when applied directly after forming the pat. A ball of this paste when dropped from a height of 70 centimeters will flatten very slightly and will not crack. A lump dropped from the point of the trowel will leave the surface of the latter comparatively clean. In forming the pat the cement should be thoroughly patted together with the flat of the trowel. This eliminates the air bubbles near the surface and also brings the excess of water to it. In forming the flat top, the hyperaqueous cement should be wiped off as : much as possible with the edge of the trowel, and the surface left smooth and firm. Difficulty number five is especially marked in slow-setting cements, as sometimes a slight indentation will persist for hours and in the judg- ment of an individual operator, may even not be fixed within thirty to PORTLAND CEMENT TESTING. 157 sixty minutes. However, this uncertainty can be greatly overcome if the needle is carefully applied at intervals of five, ten, fifteen, or twenty minutes according to the rapidity of set, indentations bemg made in a row. After the pat has become dry, the point where the needle ceases to penetrate is easily recognized (especially so if the surface is slightly moistened), and the time can then be calculated according to the number of previous indentations. Figs. 2 and 3 illustrate a quick and a normal setting cement, worked according to these directions. Nae QR Ww tm % SOS Souq s 58 8 ft op 4 Vesela hm b+ ®& gYSq ym ¢ SS) (0) . SS 2 8 QVSy ttt Lud VLLLLILZI LILLIA II LIL LILLIA L I LLL ILL LT LLL LL LLL LAL LLL LA Fie. 2 ; Wear ise ues Cao Q . a) GW Oks YIsTeesssyt RW Meare ile teuli® Se MSSUUME Gog d ~ BaSoSHoGS LYS h YVSOSCSS SVS x cn Se SO yor WLZLZTTLTLLILLLILLILLLLLILLLLL LLL ALL ALLL LLL LLL LLL LLL 3 Fig. 3. This method, once the details are mastered, is just as convenient and quick as a less accurate one. For research work and for cements the setting qualities of which are close to the requirements of specifications, it is especially valuable, as we have found that two pats of the same paste will compare almost eXactly, and that even different operators will not vary 10 per cent if they are careful and efficient. 158 REIBLING AND SALINGER. However, it has been our experience that, as Spalding’ states, “the rate of setting of neat paste gives but little indicaton of what the action may be with sand.” Several instances of satisfactory neat and unsatis- factory mortar tensile strength have been encountered in this laboratory, because of the more rapid setting of the cement when combined with sand. It is deemed sufficient to state here that the mortar and neat set must vary because of the differing percentage of water which is used, the difference in physical manipulation, in the air exposure, in mixing, the physical and possibly also the chemical influence of the sand. For the thorough study of the nature of some cements the determina- tion of neat and mortar setting qualities may be essential. A simple method to determine the setting time of a mortar is here suggested. The beginning of setting when sufficiently rapid appreciably to influence the briquette manipulation is characterized by sudden drying and a slight stiffening of the mortar. Ifa mixture is made as if for briquettes and the mortar then placed on a glass plate and divided into cubes with the trowel, a slight set may readily be detected when a cube, upon being crushed between the finger and thumb, feels dry, crumbles apart and offers a slight resistance to the crushing force. A harder set may be arbitrarily fixed and determined when the setting has progressed to the extent that a one inch cube dropped from a height of one foot will not crack. However skilled the operator may be, or however accurate his method, uniform results even by the same operator and on the same cement can not be insured unless the precautions described under the previous head- ings of “Effects of aération” and “The moist-air closet” are heeded. Thus, the first sample taken from near the surface of an exposed package not previously mixed, may set in twenty minutes, while succeeding sam- ples taken at a lower depth may not set for hours. When it is desired to make a series of comparative setting tests on the same cement it is advisable to remix the sample thoroughly before weighing and then store the cement in tightly stoppered, wide-mouthed bottles until it is used. The pats should be stored in-insulated compartments of the moist-air closet to avoid the influence of the heat liable to be generated by other pats during setting. SOUNDNESS. Tests for soundness, lke setting pats, should be made with a paste of the correct plasticity. If too little water is used in gauging, the cement will not adhere properly to the plate, and lack of cohesion in the cement itself may result in cracks not due to its subsequent expansion or contraction. If too much water is used, shrinkage cracks of such a nature as to be easily mistaken for evidence of unsoundness, may occur. © Tbid., 111. PORTLAND CEMENT TESTING. 159 This laboratory uses the same plasticity in its tests for soundness as for setting pats. The cement in this condition is thoroughly wet and pliable, but still stiff enough to retain its shape, therefore it meets all the require- ments of a just test. Uniformity between different testers is also secured, because 1 per cent of water more or less would so change the nature of the paste that it would be rendered either too dry or too liquid. The results obtained depend much upon the skill of the operator. Sudden changes in temperature during the steaming and boiling tests should always be avoided. Moistening the surface of the glass plate with a damp cloth before applying the paste will insure better adhesion to the plate. A ball of the paste should then be applied to this surface and patted down very vigorously into the desired shape. Vigorous patting with the flat of the trowel eliminates any interior cracks, reduces the air bubbles to a minimum and brings the excess water to the surface. For soundness, the top of the pat should be arched and the rim troweled to a thin edge as shown in fig. 4. Pats made in this manner will not warp or crack unless the cement is faulty. We have noticed that different testers interpret the results of soundness tests in different ways. Some operators will report as “unsound” a cement that shows the least trace of warping even after air exposure. Cements showing only slight incipient disintegration are often reported as “disintegrated.” In like manner “off plate” and “cracked plate” are often attributed to expansion and contraction. Such an interpretation is unjust to the manufacturer, as warping and cracking to some extent under certain conditions are not to be considered dangerous. A sound pat com- bined with a broken plate does not necessarily indicate dangerous con- traction or expansion. Hvery cement expands more or less, and in this case the adhesion between the cement and the glass is very strong. As the glass also has an expansion factor, all such cases should be reported as satisfactory if the pat itself shows no sign of cracking or warping. To insure a perfect understanding between the manufacturer, engineer and tester and to avoid unjust or misinterpreted results, specifications should include a descriptive chart of the proper standard interpretation by which the extent, significance, and importance of the various degrees of warping, cracking, disintegrating and shrinking are to be regulated. This labora- 160 REIBLING AND SALINGER. tory has adopted the standard portrayed and described by Taylor and Thompson *° in order to insure a complete comprehension in this respect. Much diversity of opinion exists regarding the rejection of a cement which fails to meet the boiling test,** but we regard such a cement as dangerous if it is to be used in works exposed to the heat of the tropical sum. . Excess of lime, coarseness of grinding, insufficient seasoning, and underburning of a cement may cause it to fail to pass the soundness test. If lime is the cause, storage may eliminate the defect, as the free lime would thus be changed to the carbonate, or slaked, and so would not cause subsequent expansion. Many engineers believe that failure to pass the hot test is not a proof of inferiority, as the cement so failing, if mixed with sand or some other ageregate, has produced durable masonry; it is also a known fact that thoroughly slaked lime paste can be added to a Portland cement mortar without injurious results. We suggest that, as is the case in determining the time of setting, some specification be devised to test the mortar mixture as well as the neat paste. TENSILE STRENGTH. The variation in the breaking strength of both neat and sand briquettes is a source of trouble to every cement tester, and despite every effort to eliminate this error, breaks continue to be variable with a persistence that makes it necessary to double or treble the number of briquettes otherwise required. We have made a thorough study of this variation and as a result have come to the conclusion that only a portion of it is due to the personal error of the operator, and that the remainder is caused by the characteristics of the cement itself. Personal error even with the most careful manipulation, may be pro- duced by (1) unavoidable variation in troweling; (2) difference in the force of the blows; (3) lack of equality in forming each layer of the briquette; (4) variation in the size and shape of the mold; (5) difference in the size and shape of the sand particles; (6) personal error in machine operation; (7) unavoidable internal strains and voids caused by the manipulation which the specifications impose (8) the impossibility of securing a perfectly homogeneous mixture; (9) variation in drying. The errors caused by the natural characteristics of the cement, and which need more extended explanation, are as follows: 1. It is obvious that it is impossible to expose the same number of cement particles to the action of the air for the same length of time in *° Concrete, Plain and Reinforced. New York (1907), 103-107. =UOC Cuts PORTLAND CEMENT TESTING. 161 each instance during troweling; those outside will be exposed more than the inner ones and the evaporation caused by contact with the air may cause setting. We would expect a greater variation from this cause in quick setting cements than in slow ones; and our experience has confirmed this. Slow setting cements give the least variation in tensile strength. 2. Another cause of variation is the tendency possessed by some cements to enclose air bubbles, thus producing irregular voids. 3. Unequal hardening of the exterior and the interior of the cement briquettes may cause differences in heat generated during setting and variable water action during submersion. This cause may also produce internal strains and voids. This variability would also be especially marked in quick setting cements. 4. Irregularity in the intermingling of the crystals during crystal- lization. In summarizing the above conditions, only errors which are unavoidable and such as might occur in a batch of four briquettes made and manipu- lated in the same manner and under the same conditions have been assumed, and our endeavor has therefore been, if possible, to minimize the personal error, and to this end a new type of tamper differing from that specified by the United States Army engineers was adopted. We found it impossible to raise the specified tamper exactly one-half inch at every blow, and at the same time to apply the blow just where we wanted it. A simple, accurate, easily and quickly manipulated tamper which gives the same force to every blow, and hits the exact spot desired, was therefore devised by us. A (fig. 5) is a thin, hollow cylinder open at d and closed at e. It weighs about 60 grams. B (fig. 7) is a solid brass rod which weighs just 1 pound. The end bearing the lug 6 is inserted into the cylinder A, 6 following the groove a. To manipulate this instrument, the rod B is held near the top with the thumb and forefinger of the right hand, A being held in the same way with the left. The lug 6} is drawn hard against the angle in the groove a, and the end e is placed on the surface of the cement just where it is desired to have the blow strike. The rod is then dropped and at the same time the hold on A is loosened. A little practice will enable any one to operate this tamper very quickly, and at the same time to deliver an unvarying blow due to the half-inch drop of the one pound rod. The blows can be directed at will and it is not possible to hit the edges of the mold. The United States Army specifications direct the tester to raise the tamper one-half inch above the surface of the cement. As the paste and mortar are put into the molds in a lumpy condition, no plane surface line is presented, and as we wished to control the force of each blow as much as possible, a surfacer was devised to enable us to have uniform plane. 162 REIBLING AND SALINGER. Fie. 8. a Fie. 5. PORTLAND CEMENT TESTING. 163 CUT TT Fie. 9. Bic. 10. This surfacer is made of steel and of the form shown in figs. 9 and 10. The flat surface c fits loosely into the mold. The layer of cement is placed into the latter, distributed as evenly as possible with the fingers, and then lightly pressed together with the surfacer. ‘Treating each layer in this manner also keeps the material from sliding and working around during tamping. To secure uniform effects of tamping it is also essential that the suc- cessive layers of each briquette be made as nearly equal as possible. This is easily attained by the use of a small beaker as a measure. After selecting a beaker of the correct capacity it is scooped full of cement, the excess shaken off, and the remainder turned into the mold. It is not advisable in mortar manipulation to use a measure; the mortar adheres to the glass to some extent and, in dumping, the sand readily falls out, but some cement paste remains attached to the beaker. thus changing the ratio of 1 to 3. We quickly form the mortar into a 164 REIBLING AND SALINGER. flat square on the slab, and by pressure with the edge of the trowel rapidly divide it into sixteen cubes. One such cube forms each layer of the briquettes. In tamping the last layer, it is advisable first to lay an empty mold exactly over the other. The empty mold acts as a guide for the tamper and so avoids the possibility of the loss of the full effect of a blow caused by striking the edge of the mold. The United States Army specifications direct that each layer of cement in the molds be uniformly tamped with thirty blows. There is no possible way to avoid unequal overlapping of blows with the tamper specified (both round and square). As a result, and also because of the fact that the cement is put into the molds in a more or less lumpy condition, certain voids and excess in the consistency and compactness of the resulting briquettes are unavoidable. Air spaces also form with more or less irregularity. These produce internal strains and variation in cohesion, and consequently differences in the breaks. This illustrates one case of a specification which imposes variability of results upon the tester. The American Society method eliminates the greater part of this trouble. The paste is more homogeneous and plastic (not lumpy) ; it is readily pressed into the molds by the fingers and a subsequent patting of the briquette with the flat side of the trowel will eliminate any varia- tion in compactness caused by unequal pressure of the fingers. The natural tendency in tamping briquettes is to strike the middle, narrow section more than the wider ones; it follows that the resulting briquette is denser in the middle portion. This is the main cause of bad breaks, besides giving a higher result than is just if uniformity of tamping is followed. It is just as essential not to weaken the middle section below the average density. Such a method of tamping will give good breaks, but lower the tensile strength. After experimenting with many methods to secure as uniform tamping as possible, conducive to good breaks and greatest strength, we have adopted the following method which can be accurately carried out with our automatic tamper. Fie. 11. PORTLAND CEMENT TESTING. 165 Hie. 12. The fourteen blows illustrated by fig. 11 are repeated and the final two struck directly across the middle as shown by fig. 12. Neat briquettes made in this way always break across the center in the Fairbanks roller clips, and seldom vary more than 10 per cent from the highest (5 per cent from the mean). At times, batch after batch will break within a few pounds. Again, at rarer intervals, an occasional break occurs which is 20 per cent or more away from the normal. This variation depends to a great extent upon the nature of the cement and the consistency which the per cent of water used produces. Quick- setting cements give the greatest variation in results. Sand briquettes still continue to differ considerably, as is true with all other methods. ‘he variation in the size and shape of the sand particles and the corresponding voids and excesses of cement are such that it does not seem possible to contrive any method to eliminate the differences in the tensile strength. The chief value of our system in this re- spect is that it imsures good breaks and hence gives more data to report from. For the purposes of investigation and for work which is under dispute, the question of variation in the force of the blows and their application is eliminated by our method. It is also true that the mechanical tamper renders it possible to depend upon the labor of assistants. The variations in tamping having been eliminated, a smaller number of breaks will suffice. We find that four briquettes from two batches of mortar will almost invariably coyer the entire range of pos- sibilities, and show any undue variation in the quality of a series of cement samples, this is illustrated by Table XVIII which shows the uniformity of the cement very plainly, despite the small number of breaks. 71978 ——5 166 TaBLE XVIII.—NShowing the uniformity of breaks due to the method of tamping.’ REIBLING AND SALINGER. | | Fineness, specific gravity, and set.» Tensile strength in pounds per square inch. = No. i | Neaticentent? 1 cement to 3 Fineness | Specific | Initi 1) Final sand. (100 | Specific | Initial) Fina : mesh). | 8T@Vity set. set. | i S areHyO. = | 1 day. | 7 days. |28 days.) 7 days. | 28 days. | | | | | aaa | | h. m.| he m. | | F2-1 95.9 3.06} 1 10} 2 15 345 558 625 166 250 F2-2 96.3 3.07 LG | 2 a0; 331 | 593 650 164 271 F2-3 96.6 3.07 | 1 30 2 20 | 327 | 495 | 582 162 274 F2-4 97.0| 3.07 | 1 80) 2 25) 800 fo Gab |) = es 166 260 F2-5 96.8 | 3.07 | 1 30] 2 30 322 | 520 Bib) 159 249 F2-6 96.0 3.07 | 1 23 | 233 330 528 622 157 245 F2-7 97.0 3.06 } 1 35 | 2 45 851 584 | 562 163 247 F2-8 96.5 3.07 | 1 37] 2 40 334. 506 611 163 | 260 F2-9 96.5 3.06 | 1 25 | 2 45 335 | 503 620 158 249 F2-10 96.4 3.07 1 30| 2 50 338 | 522 600 175 256 Number of briquettes broken 2] 3 3 | The following diagram demonstrates the value of our method: Twenty- eight briquettes of cement, ground extremely fine, were made and four briquettes were broken every other day. aSee also Table XI. »Soundness satisfactory at the end of six and twenty-eight days, respectively. curve shown by diagram number This curve is of interest as it plainly demonstrates the unequal action of water upon the briquettes, the tendency being for the curves of the extreme breaks gradually to grow farther apart. The low tensile strength 9 oO. Age in days Dracram No. 3. The results are plotted on the neat, as a characteristic of extremely fine grinding, is also of interest. PORTLAND CEMENT TESTING. 167 In this laboratory all sand briquettes are broken in a German machine (Hugershoff), invented by Michaelis. The Fairbank’s roller grips are so heavy and the surface of contact so narrow that they crush through the majority of %-day mortar briquettes, giving bad breaks and figures representing low tensile strength, and this is especially true of cement which does not in itself develop great strength. The German machine offers a wider surface of contact and the grips support their own weight. Comparative tests carried on for months in routine work give 10 per cent higher results with the latter, but the variation is greater, as the machine is more delicate; the probability of obtaining bad breaks is also greater, but when our system of tamping is used this probability is reduced to a minimum. Tt is especially difficult in this climate to obtain uniformity in the demonstration of tensile strength. The laboratory temperature seldom falls below 26°, and is often as high as 31°.5. Our own experiments bear out the conclusions derived from all published data on the influence of temperature. High temperature is conducive to slightly greater tensile strength on 7- and 28-day tests, and also to a greater variation between breaks. The different tensile strengths secured by different machines, molds, and grips is another reason why there is such a great lack of uniformity between different laboratories. Johnson, Sabin, Thompson and Taylor, Spalding, Butler, and in fact almost every authority on cement testing, devote considerable space to illustrating the variable results that occur from this source. A specification that allows any form of grip and mold can not hope to accomplish its purpose. The Army specifications allow the use of any tensile strength system. The American Society specifica- tions recommend a special form of briquette and regulate certain im- portant factors in the grips. To insist upon a certain machine, grip, and mold would be a rather delicate undertaking, but until this is done there may always be a large difference due to “personal equation” between the tensile strength determinations between different laboratories. The American Society introduces a very good check upon the mixing and molding of briquettes by specifying that they should be weighed just before immersion and that all which vary more than 3 per cent from the average, should be rejected; in this way greater certainty in results is obtained. It is very easy to work within these limits, and every tester should strive to attain weights which approach each other within 1 per cent. This determination of weight, in addition to being a check upon the uniformity of mixing and molding, may also disclose the effect of unequal drying and of imperfect molds. Sand briquettes are more liable to variation beyond the limits of 2 per cent than are neat. This dif- ference is due, as is the variation in breaking strength, to the lack of uniformity in the size and shape of the sand particles and the irregularity in yoids. 168 REIBLING AND SALINGER. The conclusion would naturally be drawn that the greater the density of the briquette, the greater would be its tensile strength. This is not true within limits of 2 per cent, as the other reasons for “personal error” above described may overcome the natural tendency to high tensile strength caused by the density of the material. In sand mixtures, also, a high density may simply show that more sand and less cement have been used. These facts are illustrated in our routine work and shown by Tables XIX and XX. - TABLE XIX.—WMortar briquettes, 1 to 3; 123 per cent water; tamped. ensue, densile Sample No. eae ponds eee | Sample No. es inoue ee ~ | per square i * | per square = inch. | inch. 15 ees ri 171 131.0 | HiEss= Se 7 191 131.1 7 184 129.9 | 7 177 130.7 28 214 | 130.5 | 28 221 130.0 28 200 | 126.5 | 28 233 129.2 H}l-Gyssseas 7 176 TRH I Tae) Ss 7 188 131.1 7 192 | 130.6 7 177 130.5 28 188 | 131.0 28 244 131.0 28 195 130.0 28 232 132.0 ay key eee casi 7 178 28 074 || es =9 eee 7 182 130. 0 7 191 129.3 | 7 192 130.3 28 223 128.6 || 28 238 130.8 28 216 127.9 || 28 235 131.0 TABLE XX.—Neat briquettes made from sample Y1 according to American Society specifications. s | j Tensile | Tensile 2 strength . . Average ae strength . . Average pee in in rounds Welehiaa weight in | abeaD in pounds Welghvin | Reena BYS: persquare| & : grams. YS. | persquare| ST®™s. grams. inch. inch. | | 2S Sarees | 603 137.3 636 136.5 | PIE ertes | 665 138.8 613 136.6 | 137.8 136. 6 QBS oes (539) 138.2 | Y 599 136.5 | 28 esas tee 638 137.0 | 609 136.8 | OR weee tus (686) 137.5 | 621 187.2 | 28 ae 636 137.8 | | (661) 137.7 137.9 | Ne gous linea 620 137.8 | : 631 137.0 137-0 28s se eS 625 138. 6 617 136. 4 | For some time this laboratory was forced to manipulate all cements strictly according to United States Army Engineer specifications, with 20 and 12.5 per cent of water for the neat and mortar tests, respectively. Twenty per cent of water will not satisfy the chemical and physical possibilities of many good Portland cements, and the following table illustrates this fact: PORTLAND CEMENT TESTING. 169 TaBLE XXI.—Variations in tensile strength with varying quantities of water; 7-day results. Neat. 1to3 ——| mortar, Sample No. 20 per | 22.5 per Wapigee cent cent cent water. water. water. 219 633 213 182 614 208 167 612 216 152 600 200 178 603 209 The highest results of four good breaks are recorded in each instance. It will be noticed that the sand briquettes (12.5 per cent water equal to 50 per cent calculated on the cement) present higher results than the neat with 20 per cent of water; and that 2.5 per cent additional for the paste increases its tensile strength over 200 per cent. The following table shows a failure in either case : TaBLe XXII.—Varying quantities of water used with a failing cement. 7 days neat. 28 days neat. | 5 | Sample\No: | 20per | 24per | 20per | 24 per cent cent cent cent f water. water. | water. water. | | TLS leo oe SOU ae 200 | 404 | 274 468 WU184 See es 257 438 | 350 523 TSG aes See ecences 262 418 318 451 Twenty-seven per cent of water gave lower results than 24 per cent. It is a simple matter to judge when a cement contains enough water if the method of tamping is used. The surface must be wet when the last layer has been tamped into the mold and of not quite the plasticity described for the pats used in determining the time of setting. A dry surface is positive proof that very low tensile strength will result. The determination of the “normal consistency” can not be used for this purpose as the resulting paste is too slushy for tamping. If 20 per cent of water gives too dry a mixture, we add an additional quantity sufficient to bring the water to the surface after tamping. The percentage of water necessary to accomplish this result is included with the report of the tests. The results obtained in this way by our tamper and method of tamping are satisfactory, consistent and true to the quality of the cement. The best result of four good breaks is sufficient for all routine work. The United States Army specifications state that the best results are 170 REIBLING AND SALINGER. obtained with a mortar containing 10 to 12.5 per cent of water, and suggest the use of 12.5 per cent.. This is contrary to best practice and results. The correct amount of water for sand, as for neat briquettes, depends upon the nature of the cement, and the amount of water necessary to wet the surface of the sand. We find that 12.5 per cent is too much for mortars the neat cement of which worked with 20 per cent makes a fairly wet paste; and that 10 per cent for such a cement gives better results. The reason for this is a physical one, as in tamping a very wet mortar into place, much of the cement is unavoidably lost to the briquettes. During the tamping operation the water is forced to both surfaces, and carries with it the finest (most valuable) cement particles. In finishing the briquette, this top surface, especially rich in cement, is struck off and the resulting briquette is weakened by the reduction of the 1 to 3 ratio as well as by the loss of a portion of its most valuable constituent. We give this explanation as the reason why many briquette machines fail’? and why under certain conditions a slight finger pressure will make a stronger briquette than powerful mechanical force. The con- siderable pressure exerted on the briquettes by such machines forces the water to the surface and this carries cement with it, while the sand is left in the mold. & As the addition or subtraction of as little as 1 per cent of water may effect the resulting strength of a mortar briquette sufficiently to cause the acceptance or rejection of the material, the American Society intro- duces a good feature in cement testing to cover this effect, for in their specifications the amount of water necessary for any mortar is given according to the percentage of water required to reduce the neat cement to the normal consistency paste. This is shown by the following table: Taste XXIII.—Percentage of water required for standard sand mortars. 1 part cement Normal con-| to 8 parts sistency, neat.| standard Ot- tawa sand. Per cent. Per cent. 22 Che 23 9.8 24 10.0 25 | 10.2 26 10.3 27 10.5 28 10.7 29 10.8 30 11.0 22 Eng. News (1902), 48, 130. PORTLAND CEMENT TESTING. Wl The American Society specifications, with some modifications, will be adopted to test cement for all future Philippine construction work. Although this change has been favored by this laboratory, we do not believe that the above table, regulating the amount of water for mortar briquettes, will be advisable in this climate. A natural, sieved Philippine sand will also be used, but the ratio between the results obtained with this and those with standard Ottawa sand is still to be determined. Atmospheric influences wiil not. affect the cement during mixing and molding according to these specifications to as great an extent as with the tamping method, as the whole operation of making the briquettes, once the normal consistency has been ascertained, requires only about one-third of the time. e However, the tamping method, according to the United States Army specifications, is more in accordance with actual practice. It takes from sixteen to eighteen minutes to gauge the molds, which is about the average time that concrete manipulation in structural work requires. If the cement begins to set in ten or fifteen minutes, the tensile strength of the briquettes will be reduced by subsequent tamping, which is just what may be expected to happen in field work. According to the American Society manipulation, the briquettes are gauged in five or six minutes, hence the result of quick setting ten or fifteen minutes after the water is added does not affect the tensile strength so much, as the intermingling of crystals which are then formed are not broken up by subsequent tamping. Therefore, failure to pass the initial set require- ments of cements tested according to the American Society specifications must be given more important consideration than otherwise, as the tensile strength, while little affected in laboratory tests, may suffer considerably thereby in construction work. SPECIFIC GRAVITY AND LOSS ON IGNITION. Much diversity of opinion exists among cement workers regarding the value of the specific-gravity test. It was formerly considered as an almost infallible indicator of adulteration and underburning. The work of Butler,?? Meade,?* and of the committee on technical research of the Association of Cement Manufacturers has proved that low specific gravity is often due to seasoning, and that Portland cement can be heavily adulterated and still retain a specific gravity above 3.10. As a result, many engineers do not now attribute any value whatever to this test. However, the experience of this laboratory induces us to support the 3Chem. Hng. (1907), 5, 219. “Chem. Bng. 6, 17. 172 REIBLING AND SALINGER. assertion of the paragraph headed “General observations” of the “com- mittee on standard specifications for cement.” This committee states: “Specific gravity is useful in detecting adulterations and underburning. The results of tests of specifie gravity are not necessarily conclusive as an indicator of the quality of a cement, but when in combination with the results of other 93 95 tests may afford valuable indications.” * (Italics are supplied.) The specific gravity is wseful in detecting adulterations because certain adulterations will alter the specific gravity beyond the limits of specifica- tions. However, the adulteration of Portland cement is so readily de- tected by competent chemists and testers that it is now seldom indulged in by manufacturers. The real problem of cement testing concerns itself with the pure product; and for the valuation of this we find the specific- gravity determination to be a great aid. Of course, its importance is limited. Like the chemical analysis, it gives definite aid only to a limited degree. Chemical analysis will not show the degree of burning nor the compounds that exist in a cement; and the specific gravity will not always disclose adulteration or underburning. However, both these tests give valuable aid in tracing causes of defects which by other tests have been found to exist. For instance, it was the relation between the specific gravity, the tensile strength and the setting time of the cement recorded in diagrams 1 and 2 which gave us the first clue to the cause producing the variations which prevailed throughout these tests and which led us more fully to investigate the effects of aération on high alumina cements. Now that we understand the nature of this cement, the specific-gravity determination alone enables us to predict very ac- curately what the results of the other tests will be and to suggest how the cement may be improved. Failure to pass the soundness tests may be due to two causes—excess of lime or underburning. Unsoundness in conjunction with low specific gravity proves that underburning alone is the cause of the warping and disintegrating. Cements may attain a low specific gravity as a result of prolonged seasoning. If this benefits the cement, well and good; but if it injures it, then the material should not be allowed to season, or, if seasoning has already developed dangerous properties, it should be rejected. The specific gravity, before and after ignition, will ‘indicate to what extent seasoning has effected a well-burned cement, and a record of tests com- pared with the corresponding specific gravities will show the quality of the cement developed by the absorption of various amounts of water and carbonic acid. If a cement shows little change in its specific gravity before and after ignition, and also gives unsatisfactory tests in tensile strength and setting *° Meade, loc. cit. PORTLAND CEMENT TESTING. 173 properties, chemical examination will usually show that it has not the proper “hydraulic index.” i Underburned cements usually have a very low specific gravity, because they absorb water and carbon dioxide more rapidly than well-burned cements and because all the carbonic acid may not be driven off during the burning of the raw material. The compounds formed by underburn- ing are not as stable as those of a well-burned cement and hence are more readily influenced by atmospheric conditions. Underburning is readily detected by the soundness test provided the cement is “fresh;” but seasoning often eliminates the unsoundness and therefore renders this test of no value for its detection. We must then depend upon the specific gravity, loss on ignition, color, and other tests to diselose the fact. A high loss on ignition is not characteristic of the best brand of Portland cement, even after prolonged storage. R. & W. Fresenius * consider “that the limiting value of the loss on ignition of good Portland cement should not exceed 3.4 per cent.” The following table illustrates this contention : Cement A. Cement B. Cement C. Conditions. Specific | Loss on | Specific | Loss on Specific | Loss on gravity. |ignition.| gravity. | ignition. | gravity. | ignition. : | | | eENOtiquitestnittcd=s-s=sss—ymeenan 2.92 | 3.59 3.04 1.47 2392 5.39 Shightlystritted/ === eee eee 3.105 0. 66 3.15 0,69 ese |-----==-=- Strong lyatritted esas 3.115 | 0.27 3.18 | 0.19 3.00 | 2.36 Wey Sunomvedby water |oee eee 3.19 0.24 Own ninnGl 3.05 0.65 BO O23 See Sabin“ remarks “that the determination of water and CO, may give some idea of the deterioration of a cement on storage. M. Candlot considers that in the case of Portland cement a loss on ignition (water and CO.) exceeding 3 per cent indi- cates that the cement has undergone sufficient alteration appreciably to diminish its strength. Spalding ~ affirms that “if the quantity of CO, be large, it indicates either that the burning has been incomplete or that the lime has become car- bonated by subsequent exposure. The energy of the lime is thus diminished, the portion of lime in combination with CO, being inert.” While we have not enough data to cover every instance and to formulate this as a general rule, it has been our experience that the absorption of carbonic acid and water decreases the tensile strength of every sound * Soe. Chem. Industry (1894), 13, 252. Ztschr. Anal. Chem. (1893), 32, 433, 445. 7 Sabin, Louis Carlton: Ibid., 34. 8 Toid., 4. 174 REIBLING AND SALINGER. Portland cement, even though it does not develop quick-setting properties by this exposure. Table XXIV gives a typical example: TABLE XXIV.—January 22, 1907—mortar, 1 to 3. [Specific gravity, 3.11.] | | Number 7. < ‘ — of bri- 7-day. 28-day. quettes | broken. | | ~ IAWVETAR Cisse eee See | 171 227 12 dg h estate eee res | 195 250 12 This same cement stored in coarse canvas cloth bags, twenty days and three months longer, gave the following results: (Specific gravity, 3.08 after 3 months.] Number 5 : ¢ 7 of bri- 7-day. | 28-day. quettes broken. After 20 days: Average __ = 147 212 12 Highest 2-2 os 165 280 12 After 3 months: AVClage soe Sa es 135 200 16 Highest sansa 146 211 16 From the nature of things this loss in tensile strength is not difficult to explain. It is generally understood that all cements are improved by storage, but it has been proved that this is only true of those cements which are either too high in lime or underburned. Aération renders part of the excess or free lime inert because of the formation of the carbonate of calcium and also slakes some of it by the absorption of water. Thus, the cause of unsoundness is removed in time, and the cement is gradually improved in this respect. But “the higher in lime a cement is the greater its strength is known to be if thoroughly burned,” ?® and “the maximum of lime is usually controlled- by the ° Therefore, 1f a cement is sound it does not contain excess or free lime and the carbonization of the hme in a sound cement should reduce its tensile strength, as it lowers the percentage of active lime, the carbonate of calcium being inert. With all due respect for the great value of Meade’s work, we take exception to a portion of his assertions relative to the specific gravity. soundness tests.” * ° Eng. News (1905), 53, 84. "Chem. Eng. (1907),-5, 343. PORTLAND CEMENT TESTING. 175 Among other conclusions he states “* “that low specific gravity is usually caused by seasoning of the cement or of the clinker, either of which improves the product. * * * Underburned cement is readily and promptly detected by the soundness tests and no others are needed for this purpose. * * * That the requirements of specific gravity should be omitted.” Underburning is readily detected by the soundness test, only when the cement is fresh. Seasoning of underburned cement may eliminate the causes of its unsoundness. Meade himself states in this same refer- ence that an underburned cement which, when freshly made, failed to stand a 5-hour steam test without complete disintegration, after one month’s seasoning stood 5-hour steam and boiling tests perfectly. The greater part of the cement received at this laboratory for commercial testing has been seasoned for a greater or less length of time, therefore the soundness tests are not liable to detect underburning in most instances. Cement raw material, high in alumina, fuses so readily that it is difficult to control its burning, and as a result almost all high alumina cements vary considerably. It is also very difficult to detect the relative degrees of -burning which the commercial, high-alumina cements have undergone and it is only possible to do so by taking into consideration many of the physical properties of the material. It has been observed that a brown shade,** a low specific gravity,** a high loss on ignition, the presence of blotches ** between the soundness pat and the glass plate, a high, insoluble residue and a generally erratic behavior of a cement, - exist simultaneously with a relative increase in the rate of carbonic acid and water absorption. These are all regarded as signs of underburning, and a study of all of them gives the only indications of the relative degree of burning of seasoned, high-alumina cements that we have been able to recognize. Meade’s statement that seasoning of the clinker improves a cement is also open to discussion. Some cements are improved by this procedure, but many others are not. Instances are on record where seasoning induced quick setting and low tensile strength, even when calcium sulphate was present. Meade admits “that cements should contain at least 2.5 times as much silica as alumina. Cements containing less than this amount of silica are apt to be quick setting, or else to become quick setting on exposure to air.” : It is hardly necessary to state that we do not think that the require- ments of specific gravity should be omitted from specifications. This 1 7bid., 65, 19: ® Sabin, Louis Carlton: Jbid., 36. * Soc. Chem. Industry (1894), 13, 255. “Taylor and Thompson: Jbid., 101-107. 176 REIBLING AND SALINGER. test is of great value under certain conditions. Every good Portland cement will meet its requirements before or after ignition, and therefore its determination imposes no unjust or partial test. Determination of the specific gravity will be valueless unless the effects of aération are guarded against, as the exposure of the small quantity of cement necessary for this test enables the action of the atmosphere to alter its composition very much in a short time and so to reduce its specific gravity accordingly. It is the practice of this laboratory to take the cement for the specific gravity determination from the sample at the same time that the material for the other tests is taken. It is then dried at 110° for thirty minutes and immediately put into small, glass bottles which are tightly corked until the cement has cooled ; it is only used after this procedure. The difference between the specific gravities before and after ignition indicates the amount of volatile constituents present in the cement, but when it is desired to know only the amount of carbonic acid and combined water which has been absorbed, the loss on ignition affords a much simpler and a more accurate test. CLIMATIC INFLUENCES. Local, tropical, climatic conditions must necessarily have an influence upon cement and cement testing. In the tropics, all work is done practically in the open air, being protected only from the direct rays of the sun. The climatic conditions under which cement tests or commercial work are undertaken coincide very closely with the meteorologic obser- vations which are given in the following table for the year: TABLE XXVI.—Summary of meteorologic observations taken at Manila, P. I., situated on the west coast of the Island of Luzon. Temperature. Mecee age ae aT Month, set ea aan EPR aT Ce iss (ee A aa = A iness. Mean. | Maximum. Minimum. ity. fall. | rainy days. OG! oF. °C. | oR. Cer, OF: Per ct. In. Per ct. January —_ 25 7 33.9 | 93 16.7 62 78 1.19 5 4.6 February 25.5 78 35. 6. | 96 16.1 61 74 .41 3 3.8 Narche sess 26.6 | 80 35.6 96 17.2 63 72 74 3 3.8 Aprils 26.3 93 | 37.2 99) 18.9 66 71 1.14 4 3.5 May sesso euies 26.3 83| 37.8 100 21.7 71 77 4.20 9 5.1 JU Cee 27.8 82 36.1 97 21.7 71 82 | 9. 62 16 6.8 July ses 27.2 81 35 95 21.1 70 85 | 14.57 21 7.5 August —- Ss DR 81 34.4 94 20.6 69 84 | 13. 87 20 7-5 September_____ | 26.6 80 34.4 | 94 Palaal 70 86 14.93 20 7.4 October-=2=="5 26.6 80 35 | | 95 | 20.6 69 83 | 7.54 16 6.1 November __ a 26.1 79 33. ¢ 2. 18.3 | 65 8201) 5.3) 12 5.8 December _____ 25 ids | 33.3 2 15.6 60 81 2.13 8 5.6 «The climate of Manila is hot and moist during the greater part of the year. During the spring months it is dry. The afternoon temperature of the hottest portion of the year is modified by the northeast trade winds that prevail at that season.” Brewer, Isaac W.: Personal Hygiene in Trop- ical and Semi-Tropical Countries (1908), 119. PORTLAND CEMENT TESTING. j WAT All the requirements of standard American cement specifications are based upon cement action characteristic of a colder climate. It would be possible, of course, to manipulate the cement testing itself in tropical countries at the temperature limits specified in American standards; but this could only be done at great inconvenience and at a large expense and furthermore it would not be practical, as the results of tests so conducted would not be true criteria of the behavior and value of the cement when used in construction work. The allowances and require- ments due to the effect of the relative difference in temperature between temperate and tropical chmates should therefore be taken into account in local cement specifications. During the past year this laboratory has received a number of letters upon this subject from manufacturers, engineers, contractors, testers and other cement workers. ‘These either request information or make statements regarding the influence of local climatic conditions upon various phases of cement action and manipulation. A diversity of opin- ion has been expressed in regard to the effect of these influences by men familiar with cement work, and probably this is due to the fact that Portland cement is a very variable product and therefore local conditions which would improve the quality of one brand would injure another, and vice versa, and during the past year our endeavor has been to secure a sufficient number of results with various brands of cement to throw some light on the effect produced by this climate on the tests. Careful cement testing with due consideration of all conditions is of the greatest importance in a country such as this, where much of the material comes a long distance by sea, and where the rejection of a shipment means a proportionately greater loss to the dealer, owing to the cost of transportation, and also to the engineer, as construction work may be delayed. On the other hand, construction work is very expensive in this Archipelago and therefore a rigid interpretation of specifications is necessary to provide against all possibility of the use of dangerous cement. Contrary to the general belief, the difference between local climatic conditions and those of the temperate climates exerts very little influence upon the usual standard Portland cement tests themselves. Provided the cement is of good quality the warmer temperature prevailing here usually tends to give higher results. Of course, the fineness is not affected by it, and the specific-gravity determination is made independ- ently of the surrounding temperature. The “accelerated soundness” tests especially, are benefited, as the cement does not suffer as great a change in temperature; and hence expansion and warping is not so marked. Climatic conditions improve the characteristics of early tensile strength of most cements, as the variation in temperature from day to day and from hour to hour is only slight, the temperature of the water bath is higher than in cold climates, and the temperature during gauging 178 RNIBLING AND SALINGER. is also higher; these are all factors conducive to the development of high early strength.*° Comparative tests of both sand and neat briquettes made and preserved in the cold-storage room (17° to 21° C.) and also in the laboratory (26° to 30° C.) gave almost without exception lower results, from 3 to 10 per cent, at the lower temperature. The briquettes broke more uniformly when made at the colder temperature. The differ- ence between the strength developed under both conditions was always slight and within the limits of personal error. However, the relatively high temperature of this climate will seriously effect the setting properties of some Portland cements. This is illustrated by Tables XV and XVI (pp. 152 and 153). Fortunately, the setting properties of the majority of cements are only slightly fluenced by this difference in temperature (Tables XIV and XVII). It is the experience of this laboratory that high alumina cements develop setting qualities characteristic of class 2 (Tables XV and XVI); further experimented work is necessary to determine whether this phenomenon holds true only with this class. When comparatively fresh, high-alumina cements set slowly at both temperatures (Table XIV), additional seasoning renders them slow setting at first at 17° to 21°, but quick setting at 29° to 31°, and finally quick setting at both temperatures. The development of quick setting is marked by other peculiar char- acteristics. When the absorption of carbonic acid and combined water has progressed sufficiently, no practical amount of water which can be added will retard the rapidity of setting or eliminate the early generation of much heat, but in the earlier stages of seasoning a variation of as little as 0.5 per cent of water in mixing may produce a most remarkable difference in the time of the initial and final sets. This is shown by the following table: *° TABLE XXV.—NShowing the effect of varying amounts of water on the time of setting. Sample No. Water. Condition. Paice ee \- = | Per cent. h. m.| he m. | 1 OG se eee 21 Becomes dry and noncohesiye; heats up 6° in 4 minutes_| — (*) (8) (es Rasa eee 22 Just plastic enough to mold_----------___-_-___---_____ 0 15} 1 10 TGs aes 23 Plastics’. s2eBioe = 2208 ai se A ae a ee eee Le OR e2ie2p: 5-3 ------_-- 22 JUSt plas ti cme elo ek a ee ee 20} 1 10 5:3 es see 23 IPI AStI C2 ee ra 2 a ee a a ER See 1 30) 2 25 o-0 see DONO eee 0 Sa ee eo ee ee 15 30 0-5 eae PAY = 0} - ae ee RE ee eee |e dos 220) | aImpossible. ® Annual Report Chief of Engineers, U. S. A. (1894), 234. Sabin, Louis Carlton: Cement and Concrete, New York (1905), 119-120. Alexandre, Paul: Recherches Experimentales sur les Mortiers Hydrauliques. % See also Table XVII of this paper. PORTLAND CEMENT TESTING. 179 The samples were well mixed and screened before testing; troweling was done as uniformly as possible for exactly five minutes; the atmospheric and moist- closet exposure was the same in all cases, except that there was a gradual change in temperature from 27° at 8 a. m. to 29° at 2 p. m. After troweling sample F5-1 with 21 per cent of water for about four minutes, it suddenly became hot and dry, crumbling apart. No amount of patting would cause the cement to stick together sufficiently to form a. pat. The same result was observed on repeating the operation, and a thermometer placed in the mass rose 6° in four minutes. However, upon adding 22 per cent of water to the same cement no rise in temperature was observed during troweling ; the resulting paste was sufficiently plastic to be easily molded into a pat; and the needle used for the initial set when first applied, sank about one-eighth of an inch. However, five minutes after the pat was placed in the moist-air closet, it began to heat and to dry slightly, the initial set taking place in fifteen minutes. This experiment was repeated with similar results. Twenty-three per cent of water was then used. The plastic paste, when formed into a pat, acted normally in every way and gave a satisfactory setting time. The results obtained with sample F5-3 and F5-5 were practically identical. In the case of the latter, the excess percentage of water was reduced by 0.5 per cent to determine if possible the minimum quantity necessary to effect so profound a change. ; Two important facts become evident from the above data, namely, that both the plasticity and ‘setting time of a cement, such as was being tested, are much affected after a certain quantity of water has been added by the subsequent addition of even very small amounts of the solvent.** We are not prepared to discuss fully these results at the present time, but their analogy to the phenomenon of the crystallization of certain salts from solution is striking. Many salts have a critical solution factor. Under slow evaporation they will remain in solution until a certain limiting percentage of the solvent has been reached, when the salt will crystallize almost instantly, heat being generated during the separation. A cement, the setting properties of which are so profoundly affected by the addition of eyen smal] quantities of water, may be said to have a critical solution (or hydration) point. We would hesitate to decide whether such a cement deserves to be approved. If tested accord- ing to the United States Army specifications it would fail to pass the setting test, but under those of the American Society the normal plasticity method will give it sufficient water to cause it to set slowly. An engineer in these Islands related an experience illustrating the practical importance of this problem. The mortar, after mixing, was dumped into a car and transported to its destination by rail in five minutes. Working with a large shipment of this cement no difficulty was experienced for some time, but finally when one carload reached its This same phenomenon is less delicately shown in fig. 19 of Taylor and Thompson “Concrete, Plain and Reinforced.” It will be noticed that Portland cement C (without gypsum) reached its final set even in less than thirty minutes with 20 per cent of water. With 25 per cent of water the initial set took one hour and thirty minutes and the final set five hours. ~ 180 REIBLING AND SALINGER. destination the cement had set so hard that it was removed from the car only with much difficulty. He attributed this change to the variability of the cement, but we are inclined to believe that the water added was just sufficient to bring the cement to the critical solution or hydration point and that a bucket or so of water less than was usually employed, was used in mixing, and quick setting was the result. Portland cement is most affected by local climatic conditions before and not after it is gauged. High temperature and the alternating humid and dry atmosphere are conditions under which hydration and carboniza- tion are accelerated. In consequence, the majority of commercial prod- ucts must be especially prepared to withstand tropical climates. Portland cement is very susceptible to changes under these conditions, and it is therefore essential to the best practice that cement intended for use in the tropics should develop no dangerous properties by the absorption of water and carbonic acid in normal quantities. ‘The cement problem of tropical countries depends for its solution upon the characteristics of Portland cement; and our efforts have been to determine what class of cements are least injuriously affected by exposure and seasoning. We believe that high-alumina cements are least efficient for use in tropical climates, although they have one laudable feature in that they never show the slightest inclination toward warping or disintegrating. Air, steam and boiling tests always develop perfect soundness. ‘This is probably due to the fact that aluminous raw material fuses very readily at a comparatively low heat. “Lime burned at a high heat slakes much more slowly, and is therefore more likely to be injurious than when burned at a low temperature.” *6 Aluminous cements gain most of their strength very quickly. ‘The aluminates are thought to contribute little to the final strength of the mortar, as they are not permanent compounds, but are acted upon by various salts with which they are likely to come in contact in the work. For this reason they are not adapted for work exposed to the action of air and sea water.” * “The aluminate acts in a very energetic manner upon the set, but very little upon the hardening which is caused by the silicate of lime.” Also “from the character of the silicates and the aluminates it is evident that the latter are acted upon more quickly and rapidly than the silicates, and it is to the crystallization of the lime from the aluminates that the initial set must be contributed. Subsequent hardening must be due to the liberation of lime from the silicates.” * In conformity with these quotations, it has been our experience with cements of this nature that the 7- to 28-day gain is small; that satisfactory 7-day breaks do not insure satisfactory 28-day strength; that 7-day strength may be even greater than 28-day; that little gain “Spalding, Frederick C.: [bid., 73. 8 Thid., 54. © Toid., 58. 4 Clifford, Richardson: Eng. News (1905), 53, 984. PORTLAND CEMENT TESTING. 181 in strength takes place after twenty-eight days and instances are on record where the strength of the briquettes weakened after three months. The fifty samples illustrated in diagrams numbered 1 and 2 show only an average gain (seven to twenty-eight days) by the testers, of 16.45 per cent for neat and 48.9 per cent for sand briquettes, whereas the increase desired by the Army specifications is at least 20 and 57 per cent, respectively (Table 1). “Cement giving high early strength is to be relied upon only in so far as it has been shown by experience that it is capable of maintaining such strength.” *° The fact that the early strength of this class of cement can not always be relied upon is probably due to its nonuniformity in burning. Owing to the fusibility of the calcium aluminate, which causes balling-up and sticking together in the hot zone of the kiln,*® thus preventing uniform burning, cements high in alumina are apt to be very erratic in the stability of their compounds. As a result the rapidity with which they unite with water and carbonic acid when exposed to the atmosphere varies. The relative rapidity of the absorption of carbon dioxide and water by cements under similar conditions would therefore indicate the relative degree of low burning. The most important characteristic of a high-alumina cement and the one that needs the most consideration is its susceptibility to become quick setting by exposure to the air. It has been our universal experience that Portland cements of this class containing more than 8.5 per cent of alumina always gave satisfactory results if they are tested before they have combined with more than 2 per cent of water and carbonic acid ; and that when they had combined with more than 3 per cent of volatile constituents they failed to meet the setting and tensile strength require- ments. é It would seem as if there is something radically wrong with a cement that will not withstand atmospheric exposure to such a slight extent without developing dangerous properties, and such a cement should be rejected for use, especially in this climate. A typical example, sample No. 8 as recorded in Tables VII and VIII, will suffice to illustrate this. It is difficult perhaps to realize why such a slight difference in volatile constituents should so change the quality of a cement; and that the same cement which at first set in one hour and thirty minutes (loss on igni- tion=2.63 per cent) should, after a little more aération develop such rapid setting properties, and set in twenty-three minutes (loss on igni- tion=3.92 per cent). The combination of Portland cement with water and carbonic acid absorbed from the air is represented for all purposes of discussion by the “Spalding, Frederick C.: Ibid., 88. 4% Meade: Chem. Eng. (1907), 5, 345. 719783——6* 182 REIBLING AND SALINGER. quantity of water necessary to slake the lime, if all were present as calcium oxide, and to combine with the slaked lime to form calcium carbonate, regardless of any intermediate reaction on other compounds which might be present. This change can therefore be represented by the following equations: CaO+H,0=Ca(OH), Ca(OH ),+CO,=CaCO,-+H.0 or, Ca0+CO,=CaCO, and therefore, 1 part by weight of water will unite with 3.1 parts of lime to form 4.1 parts of slaked lime, and one part of carbon dioxide will unite with 1.5 parts of lime to form 2.27 parts of calcium carbonate. From the above equations it is yery apparent how an otherwise un- sound cement is improved by the absorption of 3 or 4 per cent of water and carbonic acid. Excess of free lime causes the unsoundness and the more of this lime which is slaked or rendered inert before gauging the cement, the sounder the resulting product will be. The calcium silicates being much more stable compounds than the calcium aluminates, the latter would be acted upon first by climatic influences. The addition of lime or slaked lime to a cement retards the setting, and from the nature of the reaction, quicklime would retard the setting more than slaked lime. The natural tendency then of the lime is to off-set the quick setting properties of the aluminates. Other conditions being the same, anything which tends to reduce the activity of the lime in a slow- setting, sound cement, will increase the rate of its setting. The ignited cement of sample No. 8 had the following composition : Per cent. Silica (Si0,) 20.5 Alumina (A1,0,;) 8.6 Tron oxide (Fe,0,) 2.8 Lime (CaO) 65.4 Magnesia (MgO) 2.3 Sulphuric acid (SO,) 0.4 Table VIII shows that the cement from the bag had absorbed 0.50 per cent more carbonic acid and 0.79 per cent more combined water than that in the can. It therefore contained (equation 3) 1.13 more inert calcium carbonate and 3.24 per cent (equation 1) more slaked lime; or 0.75 per cent (equation 3) of the lime present in the raw material had been rendered inert, and 2.45 per cent had been slaked by the additional absorption of combined water and carbonic acid by the same cement stored in the bag. : The hme in combination with silica must be left out of this considera- tion as the silicates of calcium exert practically no influence upon the initial setting properties of the cement. The entire loss in active lime PORTLAND CEMENT TESTING. 183 affected the equilibrium maintained in the early setting properties by the opposing forces of the aluminates and the lime not in combination with silica. ‘Therefore, a loss in the activity of this lime representing 0.75 per cent of the total cement of sample number 8 affects this equilibrium to a degree many times greater than if the silicates would need to be taken into consideration. Synthetic experiments also show this same phenomenon. “If much more than 10 per cent alumina is present the cement is almost sure to be quick setting even with the addition of sulphates.” ** “When cement treated with sulphate of lime has regained quick set, it may again be made slow set by addition of a small quantity of lime.” * Our belief that this cement is not of good quality is also supported by universal experience. We have already stated that this class of cements gives satisfactory tests when the samples are comparatively fresh, but fails to do so after seasoning. It will be noted that the percentage of alumina and silica in sample number 8 satisfies the limits of R. K. Meade’s formula for “freshly made American Portland cements which pass standard spec- ifications for soundness, setting time, and tensile strength,” *° namely: Per cent. Silica 20-24 Alumina 5— 9 Iron oxide ; 2-4 Lime 60-63.5 l- 2 Sulphur trioxide However, they do not fall within the limits of Le Chateher’s formula for “the limits of the amount of material usually present in good com- mercial (therefore seasoned) Portland cement,” that is: Per cent. Silica 21.0-24 Alumina @ = t3 Tron oxide 24 Lime 60 —65 Magnesia 0.5— 2 Sulphur trioxide 0.6— 1.5 Water and carbonic acid 1-3 The percentage of sulphur trioxide is also lower than that given by both authors; and the loss on ignition is greater than that given by the formula which considers it. Furthermore, Meade states that “cements should contain at Icast 2.5 times as much silica as alumina. Cements containing less than this amount are apt to be quick setting or else to become quick setting on “Meade: Chem. Eng. (1907), 5, 345. ® Tbid., 349. “The Chem. Hng. (1907), 5, 349. “Trans. Am. Inst. Min. Eng. (1893). 184 REIBLING AND SALINGER. exposure to air.” Sample number 8 contained 2.38 times as much silica as alumina, and its actions supports Meade’s conclusion. Cements which contain less alumina and more silica than sample number 8 withstand exposure much better.. All of the five different cements recorded in Table XXVI below, failed in setting time and tensile strength when their seasoning had progressed as indicated by the “loss on ignition” column. However, number 5 withstood aération the best. It was only after it had stood exposed to the air for.a very long time and had united with 6.36 per cent of water and carbonic acid that it failed. TABLE X XVI. Cement 5, itue @ 3 | Constituent. Cement 1.| Cement 2.| Cement 3. | Cement 4.) Cement 5. ignited. Per cent. | Per cent. | Per cent. | Per cent. | Per cent. | Per cent. Sil can((Si On) beeen eee 20. 65 20.70 22.0 20.52 | 21.28 22.9 Alumina (1,03) —-------—-- 8.57 8.42 8.9 Serville 6.95 7.5 Tron oxide (FeO 3) ------------ 3.07 3. 01 3.0 2. 65 2,29 5) imes(Ca0) = 61,83 61. 60 59.9 61.30 61.08 65.7 Magnesia (MgO) -------------- 2.26 1.94 1.55 1.96 | 0. 21 0.2 Moisturey (i109) see aea aoe aan ee 0. 41 Of 34h Ese 0.88 0:07:23 | Sea eee Loss on ignition (water and 2.47 2.76 5.3 4.33 Gk 36s Seen enewens carbonic acid). Sulphurie acid (SO,) --------_- 0.51 OR595 | Sena ees 0. 46 1.17 1.26 Carbonic acid (COs) -------_--- 0.78 Oy4 3p ences 3. 04 4), 86)}|2 eee Seven and 28 day mortar briquettes (1 to 3), as the seasoning of the cement progressed, gave the following tests of tensile strength: 7-day. | 28-day. io tien: | | Per cent. | 936 | 320 | 2.97 | 1s7 | 247 | 4.58 | 2 || Av jo bss | Contrary to this behavior, number 4 gave the worst results and a very plastic paste made from it set in fifteen minutes with a rise in temperature from 29° to 38°.5 C. Cements numbered 1 and 2 showed only 2.47 and 2.76 per cent loss on ignition respectively and yet they were quick setting. By further investigations of this nature we hope to prove what brands of Portland cement in particular are best suited to withstand tropical climatic influences best. At present we feel justified in drawing the following conclusions as being conducive to the best results and practice for all cement operations in this and similar regions. PORTLAND CEMENT TESTING. 185 CONCLUSIONS. 1. We believe that the composition of Portland cement best adapted for use in tropical climate should be within the following limits: Per cent. Silica 22 -24 Alumina 5) = 7 Lime 62 -65 Magnesia 0.0— 4 Sulphur trioxide 1.0— 2 Water and carbonic acid 0.5— 3 2. “Soundness” in accelerated tests deserve special attention here, because of the prevailing high temperature. Perfect soundness is espe- cially important for concrete works which are exposed to the intense heat of a tropical sun. 3. “Underburning” is fatal to the efficiency of Portland cement to be used in the Tropics, as the unstable compounds so formed are most easily attacked and decomposed by the energetic atmospheric influences. 4, All “sound” cements should be protected from additional aération as much as is practicable, as otherwise quick setting or low tensile strength is liable to be developed. 5. Sound and well-burned cements, high in silica and low in alumina, will withstand climatic influences best both before and after gauging. 6. High alumina cements give fairly satisfactory results if they are used before they develop quick setting. Quick setting is sure to develop in such cements if they are exposed to the air for any considerable length of time. 7. Samples sent to the testing laboratory should be preserved in pack- ages which thoroughly protect the cement from the atmosphere. No accurate results consistent with the quality of the cement as it exists in the barrel at the time of sampling will otherwise be possible. Setting tests made at the laboratory before and after exposure should be insisted upon, and if quick setting develops by this additional seasoning the cement should be rejected. This work will be continued and our effort will be to secure samples of as many grades of cement as is possible, in order more thoroughly to test the soundness of these conclusions. EDITORIAL. PERIDINIUM. For a number of years the Bureau of Health has received many com- plaints from the residents of Bataan Province to the effect that the dumpings from the sanitary barge Pluto caused a great mortality among the fish along the shores of that province. An investigation into the matter, conducted by Deputy Commissioner H. M. Smith of the United States Fish Commission steamer Albatross, proved that the mortality among fish is in no way connected with the Pluto but is due to visitations of Peridiniwm in Manila Bay. The following is taken from a report on this subject by Dr. Smith: There have been at least three visitations of Peridiniwm in Manila Bay during the current year, a noteworthy one occurring in the. latter part of January. The discoloration of the water at that time was ob- served about the 23d of the month, and increased in intensity until the 26th or 27th, after which it rapidly diminished and practically disap- peared from the head of the bay by the 31st. Another visitation was observed during the third week in March but was less extensive than the foregoing. Whenever Peridinium has invaded Manila Bay, the water over large areas has been made turbid by minute protozoa, and at a distance has the peculiar pale reddish color characteristic of such invasions. When the water was viewed over the side of the Albatross, another color was seen; and a very pronounced iron-rust tinge was observed when the animals were closely packed. The rusty color was found to be due to contained chlorophyl. At times, dense masses of Peridiniwm floated past the Albatross in wavy bands several yards wide and hundreds of feet long. During the prevalence of these invasions, the bay is unusually phos- phorescent, and tests show that the Peridiniwm is the chief cause of the luminosity. A tumblerful of water taken at night alongside the Albatross, and found to be thick with the organisms to the exclusion of all other creatures, glowed brightly with a blue light when carried to a dark room and agitated. Whenever Peridinium has appeared in the bay, there has been a re- markable scarcity of other forms of animal life. The dense schools of small fish (Atherina and others) which are nearly always present in the surface waters of the bay, and are so conspicuous about the wharves and 187 188 EDITORIAL. vessels, disappear completely, and with them the larger fishes that prey thereon. For a number of days not a living thing of microscopic size can be seen at the surface of the water, and fish-eating birds also dis- appear. As the amount of Peridinium diminished, the small fishes grad- ually reappear in the open waters (coming either from the bottom or from places where streams enter the bay and render the water unsuitable for the protozoin). The gulls and terns also return. Finally, when the creatures have practically withdrawn, the small fish reappear in myriads. A small, salt-water aquarium on the Albatross, containing a number of different kinds of fishes and mollusks from points south of Manila, was in a very flourishing condition when the ship entered the bay one morning several weeks ago, but the same night nearly all the fishes and mollusks were killed, and examination showed myriads of the Peridinvum on the gills, etc. The few fishes that survived were rapidly succumbing, until the water supplying the aquarium was strained through a fine- meshed bolting cloth, thus eliminating the injurious organisms. Since then the fishes have been quite healthy. During the prevalence of this pest, the Manila markets contain much less fish than normally, and many stalls are entirely vacant. Inquiries among the fishermen show that there is a decided falling off in the catch and that some dead fish are reported in the baclods. The injury done to the fish, however, appears to be much less than might be expected, the known mortality among aquatic creatures being so small as to afford a noteworthy contrast to the ravages of Peridiniwm in America and Japan. AN ACCOUNT OF A HUMAN SACRIFICE HELD BY THE BAGOBOS, DISTRICT OF DAVAO, MINDANAO, P. I. A geological reconnaissance of the Island of Mindanao and the Sulu Group was begun by the division of mines of this Bureau in September, 1907. The scientific work was under the direction of W. D. Smith; the military escort which was necessary throughout most of the work was commanded by Lieut. Charles 8. Caffery, Second Infantry, United States Army. One part of this work consisted in an expedition from Kotabato to Davao, a distance of over 200 miles, 90 of which were covered by a sternwheel boat plying up the Rio Grande, or Pulangi, River, from where the party traveled overland across the Matutan and Apo Ranges to Davao Gulf. This party formed the second expedition of white men to make this entire trip. The region west of the divide is mhabited by Moros, Mohammedan tribes in a semipacified state, and Manobos and several other pagan groups live in the region on the east of the divide. Several tribes or subtribes are to be found on the slopes of Mount Apo, among which may be men- EDITORIAL. 189 tioned the Atas, Giangas, Bagobos and the Kalagans. Several Amer- icans and Spaniards have visited the people around Davao Gulf and have studied their ethnology. The Jesuits devoted themselves for many years to missionary work in Mindanao and much of scientific value was ac- complished by this learned and able body of men. Mr. Frederick Sawyer has gleaned more or less scattered information from their “letters” which he has included in his book “The Inhabitants of the Philippines,” + in which he merely refers to human sacrifices without giving any of the details, and these references are to sacrifices held only among the Giangas and Tagakaolos. Blumentritt” says even less about the Bagobos, and furthermore, he never saw any of the people of the Philippines about whom he wrote. No work has yet been carried on among these peoples by the division of ethnology of this Bureau and as it may be some time before any attempt will be made to study them, I have obtained permis- sion from the chief of that division to contribute some interesting data regarding some of their customs. We encountered Bagobos along the route for several days after we reached the Matutan Range and some of them made the trip into Davao with us; when we made the ascent of Mount Apo we spent several nights in their villages and used the people for guides and carriers. The large man in the center of the group, shown by Plate I, is Tongkaling, the chief of all these people, surrounded by some of his dependents. Tongkaling is a headman and wears the badge given him by the authorities of the Moro Province. Plate II is a view of the chief’s house. Although the Bagobos wage petty wars among themselves, they have caused little trouble for Americans. Indeed, many of those nearer the coast work on the Ameri- can plantations and do fairly well. The men of this tribe present a better appearance than do the women, and in physique and features they surpass most of the other natives in the Archipelago whom I have seen, and I have seen many of the tribes. It is said that, like the ancient Spartans, they strangle at birth all deformed children. Their hempen garments are highly decorated with shell orna- ments and with Italian beads which they procure from the Chinese. They mark with some sort of design nearly every article they use, as can be seen by examining the old chief’s shield and spear. The men are greatly addicted to the practice of tattooing; the women are not tattooed to any extent, but wear brass rings on their fingers, ears, necks, toes and ankles.* The agong, shown in the upper left-hand corner of Plate.I, is known and used all over the Malay region. I have seen one man play on as many as six of these at a time. It is the chief musical instrument in *Sawyer, F. H.: Inhabitants of the Philippines, Charles Scribner’s Sons, New York (1900), 353. 2 Globus (1882), 42, 219-222; Globus (1897), 71, 19-20. * Anyone traveling in the Bagobo country will do well to lay in a stock of beads, brass wire and cheap jewelry. 190 EDITORIAL. these districts, is made of brass and is imported from Singapore. Buyo is composed of the fruit of the betel-nut palm, locally known as bonga (Areca catechw Linn.), the fresh leaves of Piper betle Linn., and lime, to which tobacco is sometimes added. It is extensively chewed by the natives of India and Malaya. SA small, brown pigeon, of the genus Phapitroron. EDITORIAL. : 193 with it. This practice is very common among the pagans and Moros here. A converted Bagobo, named Anas, gave the writer a “bongat,’ the possession of which caused the former to be greatly feared while he was a pagan. In a fourth letter, dated Davao, July 26, 1886, the following informa- tion is given: The writer cites the case of one Maglandao (not a slave), who obtained a pair of earrings for which he could not pay; whereupon he agreed with the owner to work out the price, which was about 10 pesos. Some days later the owner of the earrings grew angry with him over some trivial matter and shot him, wounding him mortally. The offender was not a Bagobo, but hearing that the Bagobos were about to make a sacrifice, he sold them the dying man for fourteen cavans of rice. The purchasers were well pleased with the bargain, since they secured the victim cheap, as was also the other party to the transaction, for he had obtained sufficient rice to maintain himself for a year. The writer learned of this from a Bagobo who assisted at the sacrifice, and whom the writer baptized later. _Both pagans and Moros make a business of selling victims to Bagobos. When a certain governor of the district of Davao expressed his disgust at this practice, a Bagobo replied: “Is it not lawful to spend your money as you wish? Our slaves are the same as money to us, and we dispose of them agreeably to our pleasure and customs.” The writer holds them to be more barbarous than the Ammonites who sacrificed to Saturn; for these made sacrifices only at a certain period of the year, while the Bagobos make them continuously.’ Every rancheria has its feasts in honor of the devil every year. He is known as Busao, Manda- ragan, Darago, and by many other names. When a feast is to be held in his honor, there is a gathering in the house of the headman where all eat, drink, sing and dance very gaily; and the only objectionable feature of the occasion that one can see is the drunkenness commonly attendant on such occasions. They pass around their liquor, inviting one another to drink, and finally calling upon the master of the feast for a speech, they drink to the great Darago, promising to follow and honor him always, and like their forefathers, give him plenty of human blood to drink to secure his friendship and assistance in their wars. The inexperienced observer, who does not understand their language, sees nothing surprising in this; while he who knows something of the Bagobos will at once recognize the proof of the previous day’s sacrifice namely the branches placed in the joint of bamboo before which the master of ceremonies invokes the Darago, for these tell the story. When a contagious disease makes its appearance, or when a relative dies, they interpret this to mean that the Darago wants more victims, and immediately take steps to appease him and thereby save themselves from death. At the moment of sacrificing they say, “Aoaton mo ian dipanoe ini Manobo, timbac dipanoe co, so canac man sapi,” which means “Receive thou the blood of this slave as if it were my blood, for I have bought it to offer it to thee.” These words they pronounce while slashing the victim with their knives. As the great devil feeds continuously on human victims, these sacrifices must be numerous. The following is taken from the Historia de Mindanao y Jolo, by P. Francisco Combes, S. J., pages 63 and 64: The Bagobos, of a pure Indonesian race, are firmly planted on the smaller ridges of the southeast of Apo and have, therefore, as neighbors the Guiangas, the *This is contradicted in Governor Walker’s report. See p. 195. 194 EDITORIAL. Atas and the Calaganes. Moreover, they practice the barbarous customs of human sacrifices, are bold, warlike and given to drunkenness; almost all of them are of fine presence, for they immediately strangle deformed ones at birth. There are more than 12,000 of them, of whom in 1887 some 800 had been baptized: Montano and Schadenberg, and the Jesuit Fathers Gisbert and Doyle, have made especial studies of the Bagobos. Since the year 1886 only one report of a sacrifice has been recorded. It is referred to, but with no details, by Sawyer. Every detail of the following story was thoroughly investigated and is vouched for by Governor Walker and Captain Plattka, senior Constabulary inspector of the district, and I have been furnished signed copies of their reports by General Bliss, governor of the Moro Province, with his permission to publish the facts. The event was the offering of a human sacrifice to the god of evil. The place was Talon and the date December 9, 1907. I give Governor Walker’s report almost in its entirety, omitting only the names of the participants: In addition to a pencil report made under date of December 20, 1907, regard- ing a human sacrifice made by the Bagobos at Talon near Digos on December 9, 1907, I have the honor to submit herewith a full report of an investigation held by myself and the senior inspector of Constabulary at Davao. We left Davao on the morning of the 27th of December and arrived at Digos in the afternoon of the same day. An order was immediately sent to the Bago- bos of Talon to come down to Digos to meet us. On the morning of the 30th the entire population of Talon, men, women and children to the number of almost one hundred and fifty, arrived at Digos. They were informed that it was reported that a human sacrifice had been made at their town and that the authorities desired to know if it was so. Datu replied that it was true that a sacrifice had been held as stated and that both he and his people were ready to tell all about it, as to the best of their belief they had committed no crime but had only followed a religious custom practiced by themselves and their ancestors from time immemorial. From the statement made by Datu and his followers, it appears as follows : That the Bagobos have several gods,’ “Bacalad,” god of the spirits; “A ganmole Manobo,” god of good, and his wife, the goddess “Diuata,” “Mandarangan,” *® the god of evil (corresponding perhaps to our devil), and to whom sacrifice is made in order to appease his wrath, which is shown by misfortune, years of drought or evil befalling the tribe or its members; it is at times necessary to offer him human sacrifice so that he will allow the spirits of the deceased to rest. They say that in case a Bagobo of rank or influence dies and his widow is unable to secure another husband it is necessary for her to offer sacrifice to appease the spirit of her departed husband in order that she may secure another. In order that these sacrifices be not made too frequently it is customary for the old men of the town to gather once each year during the time when a certain constellation of seven stars, three at a right angle to the other four, are seen in the heavens to the east at 7 o’clock in the evening; this is said to occur once a year during ‘The fact that the names of the Bagobo gods as here given differ from those in quotations given above may be due to a misunderstanding of the interpreter or it may be that Bagobos in different localities have different names for their gods. ® Mandarangan is believed by the Tagakaolos to live in the crater of Apo. EDITORIAL. 195 the first part of the month of December. This constellation of stars is called by the Bagobos “Balatic” and is the sign of the sacrifice; that is, if a sacrifice is to occur, it must take place during the period when the stars are in this position. The old men meet and decide if enough misfortune has overtaken the tribe or village during the period since the last “sacrifice to render necessary another tribute to the god of evil. It is not necessary to offer a sacrifice for each evil, but when the misfortunes are considerable, a sacrifice is held to cover all. In this case it appears that two widows went to the datu and requested that he arrange a sacrifice to appease the spirits of their departed husbands who were bothering them. The datu called a meeting of the old men; there were present, besides himself, three other Bagobos, and these four decided that as there had not been a sacrifice since the great drought (about three years before), and that since that time many evils had befallen them, it would be well to offer a sacrifice. These four men were sent out to find a slave for the sacrifice, the finder becoming the chief of ceremonies. A henchman of the datu purchased from a Bagobo a Bilan slave boy named Sacum, about 8 years old, who was deaf and eross-eyed, and who had other defects of vision making him of little or no value as a laborer. This boy was originally received as a slave from a Bilan as a wedding present, when the Bagobo married the Bilan’s daughter about a year before. The henchman of the datu agreed to pay five agongs for the boy and took him to the house of a friend where arrangements were made for the sacrifice by calling on all who, for any reason had need to appease the evil spirits, to come and take part. Three days after the slave was brought to this house, the people met at Talon near the Inolia River a short distance from the house, this being the regular place of sacrifice. Among those present were sixty prominent men and twenty-two women of the tribe. (The datu whose picture is shown on Pl. I was there.) Being taken from the house, the boy Sacum was seated on the ground near the place of sacrifice. He was naked but no other preparation was made with regard to his person. Upon a platform or bench of bamboo about 2 feet high and a foot or two square was placed a small basket or receptacle made of the bark of the bunga tree, in which each person present and taking part in the sacrifice placed a piece of betel nut; over this the men placed their head handkerchiefs and over the handkerchiefs the women laid strips of the bark of the palma tree. Upon this the men laid their bolos, and spears were then stuck in the ground in a circle around the platform. Next, the datu, as chief of the sacrifice, made an oration which was about as follows: “OQ Mandarangan, chief of evil spirits and all the other spirits, come to our feast and accept our sacrifice. Let this sacrifice appease your wrath and take from us our misfortunes, granting us better times.” : After this the boy Sacum was brought forward, and placed against a small tree about 6 feet high; his hands were tied above his head and his body was tied to the tree with rattan strips at the waist and knees. A spear was then placed at his right side at a point below the right arm and above the margin of the rib. This lance was grasped by the two widows who, at a signal from the leader of the sacrifice, forced it through the child’s body, so that it came out on the other side. The spear was then immediately withdrawn and the body cut in two at the waist by bolos in the hands of two Bagobo men, after which the body was eut down and chopped into bits by the people present, each of whom was allowed to take a small portion as a memento of the occasion, the remainder of the body being buried in a hole prepared for it. It is said that the child was deaf and almost blind and that he did not realize 196 EDITORIAL. what was to happen to him until the moment he was tied, when he began to ery; and furthermore, that death was almost instantaneous, the only cry being one uttered when the spear first entered his side. Datu , a man about 60 years of age, says that in his life he has attended or officiated at fifty human sacrifices, ‘more or less, both among the Bagobos and Bilans, and that human sacrifice is also a practice among the Tagakaolos, although he has never been present at one held by that tribe. The Bagobos do not sacrifice any but old and decrepit or useless slaves captured from other tribes, but the Bilans sacrifice even their own people. Being asked if it was customary to eat any portion of the body sacrificed, my informant replied that it was not customary nor did he know of any case where such a thing had occurred. The last sacrifice previous to this was held at Talon during the year of the drought (about 1905) when a Bilan slave, an old man who was paralyzed in one arm, was sacrificed by Datu , his master. When asked if the sacrifice of an animal would not do as well as that of a human being, they said no, better to have no sacrifice at all. They appeared utterly unconscious of having com- mitted any crime, told their story with frankness, said it was a matter not talked about among their own people, but that if we wanted to know the facts they would give them to the authorities. They maintained that the offering of human sacrifices by their tribe was an old custom and as far as they knew was the only way to appease the wrath of the evil spirits, but they said if they were ordered to give the custom up they would do so even if the devil got them all. In view of the facts in this case as brought out in the investigation, it is not thought that it is a case for prosecution before the courts, but rather one for religious instruction in so far as it is possible to give it. When it is considered that only a year and a half ago these people could not be approached by a white man without taking to the brush, and that now they will come down out of the mountains to meet the officials to discuss a question of this kind, it is evident that they have great confidence in our Government. I explained to them that human sacrifices were wrong and would not be allowed by our Government, and furthermore that I could not let them off, but would write and explain everything to the provincial governor, who would decide what was to be done in the premises. These people have promised me that if I would assist them to secure a good location near the coast, they would move down from the mountains. I have promised them my assistance in the matter and I intend to try and get them down to a point near Digos in the near future. These accounts differ in minor points, but the essential details agree very well. I know of no white man who has witnessed this event. The fact that none of our party learned about the sacrifice until we had passed through the place where it took place shows how secret the whole affair was kept. The native foreman on a near-by American plantation, where we stopped for a day or two, was the principal actor in the scene. The Bagobos are, on the whole, very tractable and well disposed to- ward Americans, in spite of this primitive and bloody custom. I lived among them for several days and felt not the least anxiety. Good judg- ment and tact in dealing with them will doubtless enable the provincial officials to induce them to give up this practice even though they have made human sacrifices for many years. Warren D. SMITH. “G1OHASNOH SIH GNV NIIVWHDSNOL ‘1 3LWid ‘S$ ‘ON ‘III ‘I0A “IOS -NUHNOLr ‘TIHg] ['ao1aIyoVS NVA V 40 LNAODOY : HLuNg “Odv LNNOW 4O 3d01S NY3LSV3 S3HL NO 3SNOH S/ONIIVHSNOL ‘Il 3LV1d ‘€ ‘ON ‘III “I0A “10S ‘Nuno ‘TIHg] [‘aor1a1yovg NVINOY ¥ do INQO00V : HLINS *(SOSODVE 3YV SHALHOd SHL) Od¥ LNNOW 4O LINWONS 3H1L NO ALYVd HNO ‘Il 3LVId ‘2 ‘ON ‘IIT “IOA “IOS ‘Nunor ‘IH d] [ao1muovg NVWOY VY dO INAODDY : HLS Vou Tr SEPTEMBER, 1908 | No. 4 THE PHILIPPINE JOURNAL OF SCIENCE EDITED BY PAUL C. FREER, M. D., Pu. D. WITE THE COOPERATION OF MERTON L. MILLER, Pa. D.; GEORGE F. RICHMOND, M. S. W. D. SMITH, Pu. D.; A. J. COX, Pu. D. ~RAYMOND F. BACON, Pu. D.; CHARLES S. BANKS, M. S. H. D. GIBBS. B. S.; R. C. McGREGOR, A. B. is PUBLISHED BY THE BUREAU OF SCIENCE OF THE GOVERNMENT OF THE PHILIPPINE ISLANDS A. GENERAL SCIENCE MANI BUREAU OF PRINTING 1908 PREVIOUS PUBLICATIONS OF THE BUREAU OF GOVERNMENT \ LABORATORIES, 1No. 1, 1902, Biological Laboratory.—Preliminary Report of the Appearance in the Philippine Islands of a Disease Clinically Resembling Glanders. By R. P. Strong, M. D. No, 2, 1902, Chemical Laboratory.—The Preparation of Benzoyl-Acetyl Peroxide and Its Use as an Intestinal Antiseptic in Cholera and Dysentery. Preliminary Notes. By Paul C. Freer, M. D., Ph. D. iNo. 8, 1908, Biological Laboratory.—A Preliminary Report on Trypanosomiasis of Horses in the Philippine Islands. By W. E. Musgrave, M. D., and Norman E. Williamson’ 1No. 4, 1903, Serum Laboratory.—Preliminary Report on the Study of Rinderpest of Cattle and Garabaos in the Philippine Islands. By James W. Jobling, M. D. 1No. 5, 1903, Biological Laboratory.—Trypanosoma and Trypanosomiasis, with Special Heitees to Surra in the Philippine Islands. By W. HE. Musgave, M. D., and Moses T. Clegg 1 No. 6, 1903.—New and Noteworthy Plants, I. The American Element in the Philip- - pine Flora. By Elmer D. Merrill, Botanist. (Issued January 20, 1904.) 1No. 7, 1908, Chemical Laboratory.—The Gutta Percha and Rubber of the Philippine Islands. By Penoyer L. Sherman, jr., Ph. D. 1 No. 8, 1908.—A Dictionary of the Plant Names of the Philippine Islands. By Elmer D. Merrill, Botanist. 1No. 9, 1908, Biological and Serum Laboratories.—A Report on Hemorrhagic Septi- cemia in Animals in the Philippine Islands. By Paul G. Woolley, M. D., and J. W. Jobling, M. D. 1No. 10, 1908, Biological Laboratory.—Two Cases of a Peculiar Form of Hand Infection (Due to an Organism Resembling the Koch-Weeks Bacillus). By John R. McDill, M. D., and Wm. B. Wherry, M. D. 1No. 11, 1903, Biological Laboratory.—Entomological Division, Bulletin No. 1: Prelimi- nary Bulletin on Insects of the Cacao. (Prepared Especially for the Benefit of Farmers.) By Charles 8. Banks, Entomologist. 1No. 12, 1908, Biological Laboratory.—Report on Some Pulmonary Lesions Produced by the Bacillus of Hemorrhagic Septicemia of Carabaos. By Paul G. Woolley, M. D. No. 13, 1904, Biological Laboratory.——A Fatal Infection by a Hitherto Undescribed Chromogenic Bacterium: Bacillus Aureus Fetidus. By Maximilian Herzog, M. D. 1No. 14, 1904.—Serum Laboratory: Texas Fever in the Philippine Islands and the Far Hast. By J. W. Jobling, M. D., and Paul G. Woolley, M. D. Biological Laboratory: Entomological Division, Bulletin No. 2: The Australian Tick (Boophilus Australis Fuller) in the Philippine Islands. By Charles §. Banks, Entomologist. No. 15, 1904, Biological and Serum Laboratories.—Report on Bacillus Violaceus Ma- nile: A Pathogenic Micro-Organism. By Paul G. Woolley, M. D. 1No. 16, 1904, Biological Laboratory.—Protective Inoculation Against Asiatic Cholera: An Experimental Study. By Richard P. Strong, M. D. : No. 17, 1904.—New or Noteworthy Philippine Plants, 11. By Elmer D. Merrill, Botanist. 1 No. AS 1904, Biological Laboratory.—I. Amebas: Their Cultivation and Etiologic Significance. By W. E. Musgrave, M. D., and Moses T. Clegg. ee The Treatment of Tutestinal Amebiasis (Amcbic Dysentery) in the Tropics. By W. BE. Musgrave, M. D. No. 19, 1904, Biological Laboratory.—Some Observations on the Biology of the Cholera Spirillum. By W. B. Wherry, M. D. No. 20, 1904.— Biological Laboratory: 1. Does Latent or Dormant Plague Exist Where the Disease is Endemic? By Maximilian Herzog, M. D., and Charles B. Hare. Serum Laboratory: II. Broncho-Pneumonia of Cattle: Its Association with B. Bovisepticus. By Paul G. Woolley, M. D., and Walter Sorrell, D. V. S. III. Pinto (Pafio Blanco). By Paul G. Woolley, M. D. Chemical Laboratory: IV. Notes on Analysis of the Water from the Manila Water Supply. By Charles L. Bliss, M.S. Serum Laboratory: V. Frambesia: Its Occurrence in Natives in the Philippine Islands. By Paul G. Woolley, M. D. No. 21, 1904, Biological Laboratory—Some Questions Relating to the Virulence of Micro- Organisms with Particular Reference to Their Immunizing Powers. By Richard P. Strong, M. D. > No. 22, 1904, Bureau of Government Laboratories.—I. A Description of the New Build- ings of the Bureau of Government Laboratories. By Paul C. Freer, M. D., Ph. D. If, A Cataloene, of the Library of the Bureau of Government Laboratories. By Mary Polk, ibrarian 1No. 28, 1904, Biological Laboratory—Plague: Bacteriology, Morbid Anatomy, and Histopathology (Including a Consideration of Insects as Plague Carriers). By Maximilian erzog, No. 2h, 1904, Biological Laboratory.—Glanders: Its Diagnosis and Prevention (Together with a Report on Two Cases of Human Glanders Occurring in Manila and Some Notes on the Bacteriology and Polymorphism of Bacterium Mallei). By William B. Wherry, M. D. No. 25, 1904.2—Birds from the Islands of Romblon,’ Sibuyan, and Cresta de Gallo. By Richard ©. McGregor. No. 26, 1904, Biological Laboratory.—The Clinical and Pathological Significance of Balantidium Coli. By Richard P. Strong, M. D. Wo. 27, 1904.—A Review of the Identification of the Species Described in Blanco’s Flora de Filipinas. By Elmer D. Merrill, Botanist. No. 28, 1904.—I. The Polypodiaceez of the Philippine Islands. II. Hdible Philippine Fungi. By Edwin B. Copeland, Ph. D. No. 29, 1904.—I. New or Notewsrthy Philippine Plants, III. II. The Source of Manila Elemi. By Elmer D. Merrill, Botanist. No. 80, 1905, Chemical _Laboratory.—t. Autocalytic Decomposition of Silver Oxide. II. Hydration in Solution. By Gilbert N. Lewis, Ph. D. No. 81, 1905, Biological Laboratory.—I. Notes on a Case of Hematochyluria (Together with Some Observations on the Morphology of the Embryo Nematode, Filaria Nocturna). By William B. Wherry, M. D., and John R. McDill, M. D., Manila, P. I. II. A Search Into the Nitrate and Nitrite Content of Witte’s ‘““Peptone,”’ with Special Reference to Its Becigente nematle Demonstration of the Indol and Cholera-Red Reactions. By William B. erry, M. D. 1Qut of print. 4The first four bulletins in the ornithological series were published by the Ethnological Survey under the title “Bulletins of the Philippine Museum.’ Later ornithological publications of the Government appeared as publications of the Bureau of Government Laboratories, (Concluded on third page of cover.) Wishes esE NINE JOURN OF SCIENCE A. GENERAL SCIENCE Vou. III - SEPTEMBER, 1908 No. 4 THE TINGGIAN.* By Fay Cooper Coe. (From the Field Museum, Chicago, and the Bureau of Science, Manila, P. I.) INTRODUCTION. For several years the Field Museum of Natural History has been desirous of making a thorough investigation of the various Philippine peoples; but it was not until 1906 that money was available for this purpose. Through the generosity of Mr. Robert F. Cummings, of Chi- cago, ample funds were provided for a series of investigations to extend through four or six years. The first party to begin work under this ap- propriation reached the Islands in June, 1906, and was followed by a second in January, 1907. Influenced by the evidences of a highly developed ceremonial life found by Mr. Dean CO. Worcester, during his visits to Abra, and by the sugges- tions of Dr. Paul C. Freer and Dr. Merton L. Miller, the writer decided to make the Tinggian the initial field for work. The studies carried on have been along the lines of general ethnology. (with special emphasis placed on the material culture, social organization, customs, religion, mythology, and decorative art), language, and physical anthropology. The following article lays no claim to completeness, as the investiga- tions are still in progress; yet 1t seems advisable, from time to time, to publish such material as may be of interest to workers in other fields. +The name Tinggian is spelled throughout this paper in accordance with the form adopted by the division of ethnology of the Bureau of Science. In all native Philippine words in which the hard sound formerly represented by “ce” occurs this sound is represented by “k,” as in “Ilokos.” 74196 197 198 COLE. GEOGRAPHICAL DISTRIBUTION AND MIGRATIONS. The Tinggian culture group has its stronghold in the subprovince of Abra. To the north and west, it extends into Ilokos Sur and Norte as far as Kabittaoran (near Dingras). Manabo, to the south, on the Abra River, is the last pure Tinggian municipality; but Barit, Amtuagan, Ga- yaman, and Luluno are Tinggian mixed with Igorot from Agawa and Sagada. Villaviciosa is an Igorot settlement from Sagada and its vicin- ity; but Bulilising (near Villavieja) is strongly Tinggian. I am told that Sigay in Amburayan is largely made up of Tinggian emigrants from Abra, and that a few rancherias in Lepanto are also much influenced. In Tlokos Sur, south of Vigan, the whole non-Christian population is com- monly called 'Tinggian, and the people readily apply the name “Itneg” (the name by which the Tinggians distinguish their own people) to themselves. A careful survey, however, shows that very few true Ting- gian towns exist in that section. A small number are of mixed Tinggian and Igorot population, while the balance are Igorot, somewhat influenced. I failed to find any Tinggian towns south of Santa Lucia. North of this point are Ballasio, Nagbuquel, Vandrell, Rizal, Mision, Mambog, and Masinget. Towns of mixed population are Kadangla-an, Pila, Kolong- buyan (Sapang), and Montero. The other villages are Igorot colonies from Titipan, Sagada, and Fidilisan. Along the headwaters of the Saltan River in Balbalasang, Talalan, Sesekan, Patikian, and Salegseg, we find a people who in dress and looks are much like the Tinggians, and they are generally so classed. ‘These people claim a common ancestry with those of Linas, Gakab, Malibkum, and the Gobang group who originally came from Bolalay-yo (near Pa- tikian). There has been considerable intermarriage with the Igorots, and extensive migration into the Tinggian belt, but very little movement from Abra to this section. The Gobang group (including the villages of Bo-ok, Kapnay, Dewangen, and Kayabong), which is the least influenced of any of this region, must, I believe, be classed with the Kalinga. It seems, then, that here we are dealing with a population made up of Tinggians, Igorots, and Kalingas, but that, with the exception of dress, the Tinggian influence is insignificant. In the extreme northern end of the island, in the vicinity of Bangui and Claveria, and again along the Apayao River is a people who call themselves “Ishneg,” and who closely resemble in size, features, and color the people of Abra. Commissioner Worcester first indicated his belief that these people were wild Tinggians. With a view of gathering more data on this point, the writer visited these regions in the early months of this year. The material from this section has not been care- fully worked over; but there is much to indicate that these people are of common stock with the people of Abra. However, the separation must have taken place at a remote period, before the Tinggian received the THE TINGGIAN. 199 highly developed ceremonial life which distinguishes him from his neighbors. Because of the many differences in customs, and the space allowed this article, no attempt will be made to deal with the Apayao branch at this time. It is difficult to secure reliable information concerning the Tinggians in early and pre-Spanish times; but all the tales of migrations tell of movements from the coast country far back into the mountains, as the pressure of the “Christians” was felt. In many cases there was a return to the lower valleys from which these people are again being slowly driven by their Ilokano neighbors. They have no tales of an earlier home than Luzon; but the Apayaos have well-defined stories of having come from the Babuyanes (to the north of Luzon) settling near Pam- plona, Abulug, Nagilyan, and Aparri, and to have reached their present home since the advent of the Spaniards. The migration into Abra from the vicinity of Sagada has already been noted. A second, considerable movement took place from Balatok to the Ikmin River Valley, where the emigrants founded the towns of Danok, Amti, and Doa-angan. Tue is a settlement direct from Bal- balasang ; and the towns lower on the Buklok River have received many additions from there, also from Gina-an and Lubuagan. All of the villages on the headwaters of the Binongan have received emigrants from the Kagayan side; while Agsimao and other towns of the Tineg group are largely made up of Kalingas and Apayaos. There is an approximate population of twenty thousand in the towns properly classed as Ting- gian (Apayao excepted). PHYSIQUE, DRESS AND CUSTOMS. The center of the Tinggian belt is reached from Vigan, in Ilokos Sur, by a trip on a raft which takes a day, or on horseback along the Abra River. From Bangued as a center, the settlements radiate in all direc- tions. ‘To the north and east, they extend two and three days’ trips into the mountains. A few of the larger municipalities are in the broad valley of the Abra or its main tributaries, where with extensive fields and domesticated animals the Tinggian has not only successfully com- peted with his Ilokano neighbors, but has often surpassed them. In the mountains, his efforts have’ been more restricted; but with his ter- raced fields he has managed to bring much of the rugged country under cultivation. Even the steep mountain sides, where irrigation is impos- sible, are cleared, burned and planted to corn and mountain rice. The rivers contain quantities of small fish, eels, and shrimp, and many are the devices employed for their capture. By nature the man is a hunter; and he is poor, indeed, who does not own one or more dogs for use in the chase. In the leisure season, following the rice harvest, it is a common sight to see ten or a dozen men with their spears, nets and 200 COLE. dogs starting for the mountains, and at nightfall returning with the game swung on bamboo poles between them. The outdoor life has given the Tinggian almost a perfect physique. The average man is about 5 feet 4 inches in height. He is neither slight nor heavy; but his muscles are full and smooth, giving him the appearance of a trained athlete. The woman measures about 4 feet 8 inches, and like the man, is well and roundly developed. In both sexes, one is impressed with the strength of the features. The forehead is high and vaulted, the eyes are wide set and moderately open, the nose, higher than that of most Luzon peoples, compares with that of the Chinese, although the flat root and concave ridge is by no means uncommon. The skin varies from a light to dark reddish-brown; but here, again, the average Tinggians are readily distinguished from the other wild tribes by their lighter color. The hair is a glossy brown-black and is slightly wavy. The dress of the man is the clout and a belt in which he keeps small articles, about the waist. On special occasions he wears a long-sleeved jacket and in a few cases, trousers. The hair is worn long and is parted straight down the middle; the two strands are twisted, crossed in the back, then carried to the forehead where they are again crossed, and the ends are fastened by intertwining at each side of the head. Be pa e ScHULTZE: SOME PHILIPPINE CASSIDID. ] (PHIL. Journ. Scr., Vou. III, No. 4. SCHULTZE: SOME PHILIPPINE CASSIDID®.] [PuHIL. JouRN. Scr., Vou. III, No. 4. PLATE Il. ScHULTZE: SOME PHILIPPINE CassIpIp. ] (PHIL. JourN. Scr., Vou. III, No. 4. [jeri Wile SCHULTZE: SOME PHILIPPINE CASSIDIDA.] [PHIL. JourRN. ScI., Vou. III, No. 4. PLATE Iv. if Rs i i re [ PHIL. JouRN. Sctr., Vou. III, No. 4. SoME PHILIPPINE CASSIDID&. ] SCHULTZE: PLATE v. ue ne "IN S4Vv1d “1 ‘9lg "g ‘Oly z “old ‘PON ‘IIT “IOA “IOS “NuAOr “IIEQ] [&dIdISSVD ANIddITIHG ANOg : azETAHOg PROTHYMA SCHULTZEI, A NEW SPECIES OF PHILIPPINE CICINDELID. By WALTHER Horn. (Berlin, Germany.) 6 Pr. lucidicolus Chd. parum affinis; multo major; fronte prope antennarum insertionem viridi et inter oculos maculis 2 parvis discoli- dalibus cyaneis aut viridibus ornata, ceteris frontis partibus et vertice et pronoto (hujus totis marginibus lateralibus et solum modo hinc inde sulcis transversis viridi-cyanescentibus) cupro-aeneo- splendentibus, elytris atro-cyaneo-purpurascentibus nitentibus, ad marginem paullo cla- rioribus (parte humerali marginali interdum cyaneo-viridescente) ; labro flavo, dente juxta-mediano acuto, mediano ceterisque 4 obtusis aut deficientibus; fronte magis excavata inter oculos, paullulum angustiore (proportione magnitudinis totius corporis!), strigis juxta-orbitalibus grossioribus; prothorace coidenter longiore, parallelo aut ante basim perparum dilatato; elytris multo longioribus, parte suturali et apicale planioribus; maculis 3 albescentibus marginalibus; humerali sat magna ut in illa, 2 cereris paullo magis a margine distantibus, media paullulum obliqua (intus et posticem versus descendente). Corpore subtus ceruleo et cyaneo, episternis interdum hine inde violaceis; pedibus (femorum parte media aeneo-metallica, basi plus minusve testacea) et 4 primis antennarum articulis (3° et 4° interdum plus minusve testaceis), nigro- cyaneo-purpureis; palporum articulo ultimo nigricante, coxis 4 anterio- ribus (posticis metallicis cum apice parvo testaceo), trochanteribus flavis. Coxis posticis in disco sparsim punctato-setosis. Long. 12-14 mm. 2 66, Romblon Insula (Philippimae), a Dom. R. C. McGregor collecta. Typus No. 2049 in collectione “Bureau of Science,” Manilensis. 1 ¢@ differt a 2 6 6 labro brunnescente, marginibus lateralibus dilu- tioribus, dente mediano magno ornato; pronoto ante basim perparum angustato; elytrorum apice paullo brevius rotundato paulloque minus applanato ; femoribus rufo-testaceis, genis anguste metallicis ; tibiis prox- imaliter et penultimo palporum maxillarium articulo brunnescentibus. 1 @ Sibuyan Insula (Philippinae), a Dom. R. C. McGregor collecta. Typus No. 1965 in collectione “Bureau of Science,” Manilensis. 273 274 HORN. There is but little resemblance at first sight, between this bright species and the little Pr. lucidicollis Chd., although the characters given above seem not to be so striking (we must remember that the whole genus belongs to those that are exceedingly poor in good distinctive characters!). The beautiful, red-golden color of the front and pronotum contrasts very well with the almost black-purple (shining) elytra. The large size and parallel form are equally remarkable. The fine sculpture of the front (rougher near the eyes), is longitudinal, that of the vertex and pronotum transverse, the median line of the latter is very slightly impressed, the suleus on the base (between the free posterior margin and the posterior transverse strangulation) deep and well marked throughout. The punctures of the elytra are separated each from the other, only on the posterior third (near the suture from the middle) are they slightly confluent. Five slight, indistinct impressions are to be seen on each elytron: running down a short distance from the interior margin of the humeral spot; at the first quarter, nearer the suture than the lateral margin; anterior to the median and apical spots; and just before the apex. _ The penultimate joint of the labial palpus is slightly thickened. The ? seems to have the prothorax slightly narrowed towards the posterior strangulation. NOTES ON A COLLECTION OF BIRDS FROM SIQUIJOR, PHILIPPINE ISLANDS. By RicHarp C. McGrecor. (From the Zoélogical Section, Biological Laboratory, Bureau of Science, Manila, P. I.) Siquijor is a coral-rock island with an area of about 235 square kilometers; it lies in close proximity to the large Island of Negros, there being little more than 19 kilometers of water intervening between the two. As clearly explained by Worcester,’ its birds must have come into the island during comparatively recent times and three of these have developed into well-marked representative species, namely, Dicwum besti, Loriculus siquijorensis, and Jole siquijorensis. It is also noteworthy that none of the Megapodtide, Turnicide, Bucerotide, Picide, Dicru- ride, Sittide, Paride, or Timelude are known from Siquijor, although each of these families has representatives in adjacent islands. The list of species here recorded is derived from a collection made in Siquijor by Mr. Andres Celestino, assistant collector, Bureau of Science, in September, 1907, and in April and May, 1908. There are here listed nine species not previously known from Siquijor which with the 87 species given by Worcester and Bourns? make a total of 96, and there seems to be little probability of this number being greatly increased. LIST OF SPECIES NOW RECORDED FROM SIQUIJOR FOR THE FIRST TIME. Hacalfactoria lineata. Cacomantis merulinus. Calenas micobarica. Acanthopneuste borealis. Actitis hypoleucos. Motacilla melanope. Bubulcus coromandus. Anthus gustavi. Falco ernestt. * Proc. U. S. Nat. Mus. Wash. (1898), 20, 581. * Proc. U. S. Nat. Mus. Wash. (1898), 20, 564. bo aI or 276 M’GREGOR. LIST OF SPECIES COLLECTED.* PHASIANIDA. Excalfactoria lineata (Scopoli). One male and one female, both in adult plumage, were collected. Siquijor is a new locality for this species where it is known as “bun-tog.” Gallus gallus (Linneus). One jungle cock was collected. TRERONIDA. Osmotreron vernans (Linnzus). Two adult males of this handsome wood-dove. The eggs are pure white: two collected April 11, 1908, measure, 30 by 22.6 and 31.6 by 22.4. Two eggs collected at a somewhat later date measure, 27.5 by 21.2 and 27.3 by 21.2. The nest of this dove is a thin platform of coarse root- lets and tendrils with a few larger sticks as a foundation. Two or three dead leaves are scattered among the rootlets. The greatest diameter of the nest collected is about 200 millimeters and the greatest outside depth less than 30 mm. Osmotreron axillaris (Bonaparte) . Four specimens in fine plumage have the wings slightly longer than do specimens from more northern islands, but the colors are not dif- ferent. An egg taken from the oviduct of a female on April 15, 1908, measured 30 by 23.7. Phapitreron albifrons McGregor. Two specimens of the genus Phapitreron can not be distinguished from P. albifrons of Bohol. Muscadivores chalybura (Bonaparte). One male and one female of the imperial fruit pigeon are of this common variety of M. wnea. Myristicivora bicolor (Scopoli). One specimen of the nutmeg pigeon. PERISTERIDA. Streptopelia dussumieri (Temminck). Two specimens of this common turtle dove. Chalcophaps indica (Linneus). One specimen. * The metric system is used in all measurements here recorded. The vernacular names given under various species were found in use among the residents of the island and were collected by Mr. Celestino. bo =I “I BIRDS FROM SIQUIJOR. Caloenas nicobarica (Linnzus). One female specimen and two live examples of the handsome Nicobar pigeon were secured in Siquijor, a new locality for this widely distrib- uted species. RALLIDA. Hypoteenidia torquata (Linneus). One female in fine, adult plumage; the local name is “tic-ling.” Amaurornis phoenicura (Forster). One full-plumaged male and one slightly immature male; in the latter the breast is somewhat mottled with slaty gray. Name in Si- quijor, “qui-yao.” LARIDA. Sterna boreotis (Bangs). One male was taken on September 7, 1907. CHARADRIIDA. Squatarola squatarola (Linnezus). A male in intermediate plumage was taken September 6, 1907. Charadrius fulvus (Gmelin). A male in intermediate plumage was taken September 6, 1907. Actitis hypoleucos (Linnzus). A female was collected in September, 1907. Siquijor is thus added to the long list of islands from which this common sandpiper is known. ARDEIDAZA. Nycticorax manillensis Vigors. A fragment, consisting of head and neck, is identified as belonging to the Philippine night-heron. Bubulcus coromandus (Boddaert). A male of the cattle egret in breeding plumage. This species is ealled “ta-la-bong” in Siquijor. Not previously noted from that island. ANATIDA. Dendrocygna arcuata (Horsfield) . A male of this common tree-duck. FALCONIDAZ. Haliastur intermedius Gurney. This common buzzard is represented by the tail of an immature in- dividual. It is known in Siquijor as “ba-nég.” Falco ernesti Sharpe. A fine male falcon is identified as being of this rare species. Although in somewhat immature plumage, a number of feathers on flanks and 278 M’GREGOR. thighs, particularly on the latter, show the characteristic smoky gray color and the close-set, black bands. The measurements of this specimen follow: Wing, 300 millimeters; tail, 160; bill from front of cere, 21; tarsus, 45. The name of this species in Siquijor is “a-na--nang-quil.” BUBONIDA. Ninox philippensis Bonaparte. The single male obtained in Siquijor does not differ from specimens taken in Luzon. CACATUIDA. Cacatua hematuropygia (P. L. 8. Miiller). Two males of the common Philippine cockatoo. PSITTACIDA. Tanygnathus lucionensis (Linneus). Two males collected. Loriculus siquijorensis Steere. This distinct species is similar to L. mindorensis but the red patch on the forehead ends in a point instead of ending squarely and the red patch on the breast covers about one-half the area that it does in L. mindorensis. The native name is “co-lan-st.” ALCEDINIDZA. Alcedo bengalensis Brisson. One specimen. Halcyon gularis (Kuhl). One male specimen of this common kingfisher; it has two names in Siquijor, “wak-bd-ta” and “ma-nak-sak.” Halcyon chloris (Boddaert) . One slightly immature female. MEROPIDZA. Merops philippinus Linnzus. Two specimens. CYPSELIDA. Collocalia troglodytes Gray. One male specimen ; known as “sai-ao.”’ CUCULIDA. Cacomantis merulinus (Scopoli). LIT This common cuckoo is called “yoi-hé” in Siquijor; this is the first record of its occurrence in that island. Eudynamis mindanensis (Linnzus). The male collected shows no peculiarities. The local name is “cu- la-hao.” BIRDS FROM SIQUIJOR. 279 Centropus viridis (Scopoli) . The local name is “cuk-cuk.” The fresh eggs, taken on April 7, 1908, measure 30.6 by 25.6 and 31.1 by 25.7. They are pure white. The nest was composed of sticks and had a small entrance in one side, the entire top being covered. The nest was built in a small bush at the height of a man’s head. PITTIDA. Pitta atricapilla Lesson. Two specimens were obtaimed of this widely distributed ant-thrush. Its local name is “wao-ha.” HIRUNDINIDA. Hirundo javanica Sparrmann. One specimen. MUSCICAPIDZ. Hemichelidon griseisticta (Swinhoe) . One male specimen. Cyornis philippinensis Sharpe. One specimen ; this species is known as “ca-man-ti-gon.” Hypothymis occipitalis (Vigors). One female. Rhipidura nigritorquis Vigors. This common flycatcher is called “ba-li-d-la” in Siquijor. CAMPOPHAGIDA-. Lalage niger (Forster). Name in Siquijor “bu-ga-wng-on.” PYCNONOTIDA. lole siquijorensis Steere. This interesting species appears to be fairly abundant in Siquijor where it is known as “tig-ba-ya.” Its nearest relatives are [. monticola of Cebu and- I. cimereiceps of Tablas. It differs from either of these species in having the top of head seal-brown without ashy gray tips to the feathers. TURDIDA. Petrophila manilla Boddaert. One female specimen; known as “yi-ta yti-ta.” Copsychus mindanensis (Gmelin). One male specimen ; called “a-ni-ni-hol.” Pratincola caprata (Linneus). One female. 280 M’ GREGOR. SYLVIIDA. Cisticola exilis (Vigors and Horsfield). This common grass warbler is called “pi-rdt” in Siquijor. Acanthopneuste borealis (Blasius). This migratory warbler seems to have been unnoticed by previous collectors; two specimens were taken by Celestino. ARTAMIDA. Artamus leucorynchus (Linneus). One specimen. LANIIDA. Cephalophoneus nasutus (Scopoli). Three specimens in good plumage. Otomela lucionensis (Linneus). One specimen in immature plumage. The name “ti-ba-las” is used for both Otomela and Cephalophoneus. Hyloterpe apoensis Mearns. A male from Siquijor agrees with numerous specimens of Hyloterpe from Bohol and these are best identified with H. apoensis, at the same time it may be noted that the single male of typical apoensis before me has a slightly smaller bill than any of the Bohol specimens or the single male from Siquijor. e ZOSTEROPIDA. Zosterops siquijorensis Bourns and Worcester. 5 Numerous specimens; known as “‘Ju-lai-og”. This bird is closely related to the species found in Basilan and Bohol but in the Siquijor species the sides of the breast are much lighter gray. DICAIDA. Diczeum besti Steere. A fair series was obtained of this near relative of D. cinereigulare. Local name “pis-pis”. Diczeeum pygmzeum (\Ixittlitz). One specimen of this plain flower-pecker. NECTARINIIDA. Cinnyris sperata (Linneus). Two adult males. Cinnyris jugularis (Linneus). One female. BIRDS FROM SIQUIJOR. 281 MOTACILLIDA. Motacilla melanope Pallas. Two females in rather poor plumage serve to add one more to the islands from which this species is known. Its name in Siquijor is “a-na-nok-yod™. Anthus rufulus Vieillot. Known in Siquijor as “a-la-lak-sing”. Anthus gustavi Swinhoe. One specimen ; not previously known from Siquijor. ORIOLIDA. Oriolus chinensis Linneus. One specimen of this large oriole. STURNIDA.. Sarcops melanonotus Grant. The two bald starlings collected in Siquijor certainly approach this recently named race. Lamprocorax panayensis (Scopoli). Two specimens. CORVIDA. Corone philippina (Bonaparte). The Philippine crow concludes the list of species from Siquijor. SOME NECESSARY CHANGES IN THE NAMES OF PHILIPPINE BIRDS. By Ricwarp C. McGrecor. (From the Zoological Section, Biological Laboratory, Bureaw of Science, Manila, P. I.) N Aluco longimembris (Jerdon). Stria candida (not of Latham) TickEnn, Jour. As. Soc. Bengal (1833), 2, 572. Striz longimembris JERDON, Madras Journ. (1839), 10, 86. Dr. Charles W. Richmond writes me that the name usually applied 1o the grass owl is preoccupied by Stria candida Latham, Suppl. Ind. Orn. (1801), p. xiv which is a synonym for the snowy owl. The next available name is the one used by Jerdon. The generic name Strix Linneus is replaced by Aluco Flemng. Cf. Auk (1908), 25, 370. Megalurus tweeddalei new name. Megalurus ruficeps (not of Sykes) Twreppatr, Ann. & Mag. Nat. Hist. (1877), 20, 94; Proc. Zobl. Soc. (1877), 695, pl. 72. The name Megalurus ? ruficeps is used by: Sykes for an Indian species in Proce. Zool. Soc. (1832), 91. As this invalidates the use of this” name for the Philippine species, the latter, figured and described by Lord Tweeddale, may be known as Megalurus tweeddalei. Zosterops boholensis new name. Zosterops leta (not of De Vis) McGrercor, Phil. Jour. Sei. (1907), 2, Sec. A, 329. The name under which the silver-eye of Bohol was described had been used before for a species from New Guinea, see Zosterops lwta De Vis, Ibis (1897), 385. The Bohol species may therefore be named for the island which it inhabits. Cinnyris henkei Meyer. Cinnyris henkei Mrymr, Zeitschr. fiir Ges. Orn. (1884), 207, pl. 7. Cinnyris whiteheadi GRANT, Bull. Brit. Orn. Club (1894), 2, 1; Ibis (1894), 514, pl. 14, fig. 1. The black-backed sun-bird, described and figured by Grant as Cinnyris whitehead, finds an earlier name in Cinnyris henkei as indicated by Dubois, Syn. Ay. (1902), 699. 74196——8 283 PHILIPPINE ORNITHOLOGICAL LITERATURE, I. By Ricuarp C. MeGrecor. (I'rom the Zodlogical Division, Biological Laboratory, Bureau of Science, Manila, P. I.) This is the first of a series of papers having for its final object a complete bibliography of Philippine ornithology. Primarily, however, these papers are intended to indicate exactly what works are available in Manila and to explain the bearing of each paper or volume on the study of Philippine birds. Particular effort will be made to record, promptly, recent and current publications which deal specifically with the Philippine ornis, but older works and those of a more general character will be included. Hach installment of Philippine ornitholo- gical literature will consist of about fifty titles and will be issued from time to time as material becomes available. Gadow, H.: Paridie and Laniide (titmice and shrikes), and Certhio- morph (creepers and nuthatches). Cat. Birds Brit. Mus. London (1883), 8, 1-386, pls. 1-9. This essential volume contains synonymy, keys, and descriptions of all species of titmice, shrikes, and nuthatches known at date of publication. No Philippine species is figured. Gadow, H.: Nectariniide. Cat. Birds Brit. Mus. London (1884), 9, 1-126, pl. 1. This essential volume contains synonymy and descriptions of all species of sun-birds known at the date of publication. No Philippine species is figured. Grant, W. R. 0.: Steganopodes (cormorants, gannets, frigate-birds, tropic-birds, and pelicans). Pygopodes (divers and grebes). Cat. Birds Brit. Mus. London (1898), 26, 329-558, pls. 5a-8. This includes synonymy, keys, and descriptions of all Philippine species in the Steganopodes and Pygopodes known at date of publication. No Phil- ippine species is figured. This work is essential. Grant, W. R. 0.: Bucerotes and Trogones. Cat. Birds Brit. Mus. Lon- don (1892), 17, 347-497, pls. 13-17. Synonymy, keys, and descriptions of all the known trogons and hornbills. No plates of Philippine species. Gymnolwmus, new genus for Anthraco- ceros lempriert Sharpe. This work is essential. 285 286 M’GREGOR. Grant, W. R. 0.: Game birds (Pterocletes, Galline, Opisthocomi, Hemi- podii. Cat. Birds Brit. Mus. London (1893), 22, 1-588, pls. 1-8. Descriptions and synonymy of all Philippine species of Megapodiid@, Pha- sianide@, and Turnicide are included, but no plates of Philippine species. Grant, W. R. 0.: On the birds of the Philippine Islands——Part I. Mount Arayat, Central Luzon. Jbis (1894), 406-411. This is the first of the series of important papers by Grant on the collee- tions made by John Whitehead; 40 species are listed. Grant, W. R. 0.: On the birds of the Philippine Islands—Part Il. The highlands of north Luzon, 5,000 feet. Jbis (1894), 501-522, pls. 14 and 15. This is an important paper on a collection of 94 species from (vicinity of Trinidad?), Benguet Province. Two'new species, Muscicapula luzoniensis and Cittia seebohmi, are described and the following 17 species, first char- acterized in Bull. Brit. Orn. Club, are more fully described: Scops longicornis, Oriolus albiloris, Stoparola nigrimentalis, Merula thomassoni Seebohm, Chimarrhornis bicolor, Zosterornis whiteheadi, Hyloterpe albiventris, Lanius validirostris, Dendrophila mesoleuca, Aithopyga flavipectus, Budrepanis jeffery, Cinnyris whiteheadi, Cinnyris obscurior, Diceum luzoniense, Dicawum obscurum, Lowia luzoniensis, and Chlorura brunneiventris. Two species, | Emberiza pusilla and H. sulfurata, are recorded as new to the Philippine avifauna. Cinnyris whiteheadi, Stoparola nigrimentalis, Zosterornis white- headi, and Chimarrhornis bicolor are figured on the two plates. Grant, W. R. 0.: On the birds of the Philippine Islands.—Part ITI. The mountains of the Province of Isabela, in the extreme northeast of Luzon. Ibis (1895), 106-117, pls. 4 and 5. This important paper lists 30 species from Isabela Province and contains additional notes on Oriolus isabelle and O. albiloris, which were previously described in Bull. Brit. Orn. Club. Zosterornis striatus is deseribed as new; Munia formosana and Cotile sinensis are added to the Philippine list. Zosterornis striatus, Dendrophila mesoleuca, Hthopyga fiavipectus, and Eudrepanis jefferyi are the subjects of the two plates. Grant, W. R. 0.: On the birds of the Philippine Islands.—Part IV. The Province of Albay, southeast Luzon, and the adjacent Island of Catanduanes. Jbis (1895), 249-267. This paper lists 46 (?) species from Albay Province and 48 species from Catanduanes. Additional descriptions and notes are given of Callaeops periopthalmica, Zosterops lugonica, and Cinnyris excellens which were previously described in Bull. Brit. Orn. Club. There are extended notes on variation in Sarcops calvus and Prioniturus discurus. Carpophaga polio- cephala is recorded from Luzon for the first time and Hmberiza spodocephala from Catanduanes is recorded as new to the Philippines. Grant, W. R. 0.: On the birds of the Philippine Islands—Part V. The highlands of the Province of Lepanto, north Luzon. Ibis (1895), 433-472, pls. 12-14. Notes on 110 species. Additional descriptions of Scops longicornis, Scops whiteheadi, Rhinamyias insignis, Lusciniola seebohmi, Cettia seebohmi, Brachypteryx poliogyna, PseudothaYrhaleus caudatus, Zosterops aureiloris, Pyrrhula leucogenis, Batrachostomus microrhynchus, and Prioniturus monta- PHILIPPIND ORNITHOLOGICAL LITHPRATURE, I. 287 nus, which were previously described in Bull. Brit. Orn. Club. Collocalia whiteheads is deseribed as new; Cinnyris obscurior is rejected as being based on O. jugularis in worn plumage; Pitta kochi and Ptilopus marchei and the females of Dicewwm wanthopygium and Cittocincla luzoniensis are fully described. Notes on plumages of Macropygia tenwirostris; Falco severus added to the Luzon list; Cerchneis tinnunculus and Collocalia linchi added to the Philippine list. Grant, W. R. 0.: On the birds of the Philippine Islands.—Part VI. The vicinity of Cape Engano, northeast Luzon, Manila Bay, and Fuga Island, Babuyan Group. Jbis, (1896), 101-128, pl. 3. Additional notes and descriptions of Siphia enganensis, Hypsipetes fugen- sis, Orthotomus chloronotus, and Zosterornis dennistownt all previously described in Bull. Brit. Orn. Club. Critical notes on the plumages of Accipiter gularis. Young male of Parus semilarvatus described. First Luzon record of I’regata minor. Notes on the Hudynanis of Fuga. Ortho- tomus chloronotus and Zosterornis dennistoum are figured on the plate. Grant, W. R. 0.: On the birds of the Philippine Islands——Part VII. The highlands of Mindoro. Jbis (1896), 457-477, pl. 11. A list of 52 species with numerous notes. Ninox mindorensis and T'urdus mindorensis, new species. Oarpophaga mindorensis Whitehead, is re- described and figured. Locustella ochotensis is recorded, for the second time, from the Philippines. Notes on Scops sp. ine., later described by Whitehead as S. mindorensis. A useful key to the Philippine species of Iyngipicus is given. Grant, W. R. 0.: On the birds of the Philippine Islands—Part VIII. The highlands of Negros. Ibis (1896), 525-565. A list of 86 species; Artamides cebuensis, Turdus nigrorum, Brachypteryx brunneiceps, and Cittocincla nigrorum, new species. Keys to the Philippine species of Oriolus, Rhinomyias, Artamides, and Hdoliisoma. Oriolus basila- micus, new name. First Negros records for Tanygnathus everetti and Sur- niculus velutinus ; first Philippine record for Cuculus micropterus. Tole gui- marasensis is considered to be distinct from Jole philippensis; Munia brun- neiceps is considered to be the worn plumage of M. jagori. Notes on the plumages of Spilornis panayensis and Falco ernesti. Grant, W. R. 0.: On the birds of the Philippine Islands—Part IX. The Islands of Samar and Leyte. Ibis (1897), 209-250, pls. 5 and 6. A list of 93 species with numerous notes. Additional notes on Pithe- cophaga jefferyi, Rhabdornis minor, and Khabdornis inornatus, which were previously described in Bull. Brit. Orn. Club. Microhierax meridionalis is described as new; notes on Ceyx ; female of Microstictus fuliginosus described. Keys to the Philippine species of Orthotomus, Macronus, Zosterornis, and Rhabdornis. Poliolophus basilanicus is considered not distinct from P. urostictus. - Hight species new to Samar and 14 species new to Leyte. Grant, W. R. 0.: On the birds collected by Mr. Walter Goodfellow on the volcano of Apo and in its vicinity, in southeast Mindanao, Philippine Islands. Jbis (1906), 465-505, pls. 18 and 19. A list of 124 species with many critical and field notes. Hypocryptadius cinnamomeus, Rhinomyias goodfellowi, and Pericrocotus johnstoniew are 288 M’GREGOR. figured. This paper lists no new species but contains redescriptions of many species previously described in Bull. Brit. Orn. Club. Geocichla andromeda recorded as new to the Philippines. Four species, Caprimulgus griseatus, Turdus obscurus, Anthus maculatus, and Anthus gustavi, are noted as new to Mindanao. Grant, W. R. 0., and Whitehead, J.: On the nests and eggs of some rare Philippine birds. Jbis (1898), 231-247, pls. 5 and 6. Short descriptions and measurements of the eggs of 36 species, mostly from specimens collected by Whitehead. The two plates illustrate the eggs of sixteen species. Hargitt, E.: Scansores, containing the family Picide. Cat. Birds Brit. Mus. London (1890), 18, 1-598, pls. 1-15. Synonymy, keys, and descriptions of all the woodpeckers known at date of publication. The Philippine species figured are Chrysocolaptes rufopunc- tatus and Thriponax pectoralis. This book is essential. - Hartert, E.: Cypselide, Caprimulgide, and Podargide. Cat. Birds Brit. Mus. London (1892), 16, 484-652, pls. 10-14. Synonymy, keys, and descriptions of all swifts, night-jars, and frogmouths known at date of publication. Plates of Caprimulgus griseatus and Lyncor- nis mindanensis. Essential for a study of these families. Hartert, E.: Die bisher bekannten Vogel. von Mindoro, nebst Bemer- kungen iiber einige Vogel von anderen Inseln der Philippmen- Gruppe. Jour. fiir Orn. (1891), 87-206, 292-302. Notes on 68 species, mostly from Mindoro. The introduction contains a short account of the work done by Philippine collectors. Kutter, F.: Beitrag zur Ornis der Philippinen. Jowr. fiir Orn. (1883), 1-28 of reprint. An annotated list of 54 species collected by Koch and Schadenberg at Sibulan, southern Mindanao. Graucalus kochii is the only new species. Yungipicus maculatus is added to the Guimaras list; Collocalia linchi, Lanius nasutus, Hypothymis superciliaris, Zeocephus rufus, Dendrophila enochlamys, Oxycerca everetti, and Hxcalfactoria chinensis are recorded for the first time from Mindanao. Salvadori, T.: Catalogue of the Columb, or pigeons, in the collection of the British Museum. Cat. Birds Brit. Mus. London (1893), 21, 1-676, pls. 1-15. Contains synonymy, keys, and descriptions of all doves and pigeons known at date of publication. Osmotreron axillaris is the only Philippine species figured. Phabotreron occipitalis is described as new. This volume is neces- sary for a study of the Columb. Salvadori, T.: Catalogue of the Psittaci, or parrots, in the collection of the British Museum. Cat. Birds Brit. Mus. London (1891), 20, 1-660, pls. 1-18. ~ Contains synonymy, keys, and descriptions of all the known parrots. The Philippine species figured are: Tanygnathus everetti, Tanygnathus burbidgei, and Bolbopsittacus intermedius, the last a new species. Bolbopsittacus is a new genus with Psittacus lunulatus Scopoli as the type. This volume is quite essential for the study of the parrots. PHILIPPINE ORNITHOLOGICAL LITERATURE, I. 289 Salvadori, T.: Anseres. Cat. Birds Brit. Mus. London (1895), 27, 23-93, pls. 1-5. An important review of the ducks, geese, and swans with keys, synonymy, and descriptions. Dendrocygna guttulata is the subject of Plate I. Salvadori, T.: On a rare species of lorikeet in the Rothschild collection. Ibis (1891), 48-51, pl. 3. Description and plate of Loriculus bonapartet. Salvadori, T.: On Melaniparus semilarvatus. Ibis (1879), 300-309, fol, & Plate and notes on this species. Salvin, 0.: Tubinares (petrels and albatrosses). Cat. Birds Brit. Mus. London (1896), 25, 340-455, pls. 1-8. A review of this order with keys, synonymy, and Cesena alo: The only Philippine species is described on page 370. Saunders, H.: Gavide (terns, gulls, and skuas). Cat. Birds Brit. Mus. London (1896), 25, 1-339. Synonymy, keys, and descriptions of all known species of gulls and terns. No Philippine species is figured. Sclater, P. L.: Pittide and Hurylemide. Cat. Birds Brit. Mus. London (1888), 14, 411-449; 454-470. Synonymy, keys, and descriptions of all the pittas and rollers known at date of publication. No species from either of these families is figured. This is an important work. Seebohm, H.: Catalogue of the Passeriformes, or perching birds, in the collection of the British Museum. Cichlomorphe: Part II contain- ing the family Turdide (warblers and thrushes). Cat. Birds Brit. Mus. London (1881), 5, 1-426, pls. 1-18. Synonymy, keys, and description of all species known in this group at the date of publication. Locustella fasciolata is the only Philippine species among the plates. Sharpe, R. B.: Catalogue of the Accipitres, or diurnal birds of prey, in the collection of the British Museum. Cat. Birds Brit. Mus. London (1874), 1, 1-480, pls. 1-14. Synonymy and descriptions with keys of all species of eagles, hawks, etc., known at time of publication. The Philippine species figured are: Astur soloensis, Astur cuculoides, and Baza magnirostris. A necessary volume. Sharpe, R. B.: Catalogue of the Striges, or nocturnal birds of prey, in the collection of the British Museum. Cat. Birds Brit. Mus. London (1875), 2, 1-326, pls. 1-14. Synonymy and descriptions with keys of all species of owls known at date of publication. No Philippine species is figured. A necessary volume. 290 M’GREGOR. Sharpe, R. B.: Professor Steere’s expedition to the Philippines. Nature (1876), 14, 297, 298. A short account of Steere’s first expedition to the Philippines and brief preliminary descriptions of ten new species: Eurylemus steervi, Phyllornis palawanensis, Brachyurus steerii, Bthopyga magnifica, Hthopyga shelleyi, Athopyga pulcherrima, Arachnothera dilutior, Diceum dorsale, Dicewm hypoleucum, and Diceum haematostictum. Sharpe, R. B.: Catalogue of the Passeriformes, or perching birds, in the collection of the British Museum. Coliomorphe containing the families Corvide, Paradiseide, Oriolide, Dicruride, and Priono- pide. Cat. Birds Brit. Mus. London (1877), 3, 1-344, pls. 1-14. Synonymy, keys, and descriptions of all species of the families named in the title. Oriolus steerii new species is the only Philippine species figured. An important work. Sharpe, R. B.: Catalogue of the Passeriformes, or perching birds, in the collection of the British Museum. Cichlomorphe: part I contain- ing the families Campophagide and Muscicapide. Cat. Birds Brit. Mus. London (1879), 4, 1-494, pls. 1-14. Contains keys, synonymy, and descriptions of the species in the families named. The work is important. No Philippine species is figured. Sharpe, R. B.: A contribution to the avifauna of the Sooloo Islands. Proc. Zool. Soc. London (1879), 311-317. Notes on 20 species mostly collected by Burbidge in the Island of Sulu. The new species described are Tanygnathus burbidgii and Gallus strami- neicollis. Sharpe, R. B.: A list of the birds of Labuan Island and its dependencies. Proc. Zool. Soc. London (1879), 317-354, pl. 30. A list of 137 species largely from the collections of Low, Ussher, and Treacher, with valuable notes. The paper includes interesting references to a number of species which range to the Philippines. Sharpe, R. B.: Catalogue of the Passeriformes, or perching birds, in the collection of the British Museum. Cichlomorphx: part III contain- ing the first portion of the family Timeliide (babbling-thrushes). Cat. Birds Brit. Mus. London (1881), 6, 1-422, pls. 1-18. An important work on the first part of the Timeliidw with keys, descrip- tions, and synonymy. The Philippine species figured are: Jole rufigularis, Criniger frater, and Criniger palawanensis. Sharpe, R. B.: Catalogue of the Passeriformes, or perching birds, in the collection of the British Museum. Cichlomorphe: part TV contain- ing the concluding portion of the family Timeliide (babbling- thrushes). Cat. Birds Brit. Mus. London (1883), 7, 1-700, pls. 1-15. An important work on part of the family Timeliide with keys, synonymy, and descriptions; no Philippine species figured. PHILIPPINE ORNITHOLOGICAL LITERATURE, I. 291 Sharpe, R. B.: Description of a new species of hornbill from the Island of Palawan. Proc. Zodl. Soc. London (1885), 446, pl. 26. Description and plate of Anthracoceros lemprieri. Sharpe, R. B.: Catalogue of the Passeriformes, or perching birds, in the collection of the British Museum. Fringilliformes: part I contain- ing the families Dicwide, Hirundinide, Ampelide, Mniotiltide, and Motacillide. Cat. Birds Brit. Mus. London (1885), 10, 1-682, pls. 1-12. Synonymy, keys, and descriptions of all the flower-peckers, swallows, and wag-tails known at date of publication. Philippine species figured are Mota- cilla ocularis and Motacilla flava. A very important volume. Sharpe, R. B.: Notes on a collection of birds made by Mr. John White- head on the mountain of Kina Balu, in northern Borneo, with descriptions of new species. Jbis (1887), 435-454, pls. 13 and 14. Cryptolopha montis new species, described. Sharpe, R. B.: On a collection of birds from the Island of Palawan. Ibis (1888), 193-204, pls. 3 and 4. A list of 129 species collected by Whitehead. Eight new species are described: Prioniturus cyaneiceps, Baza leucopias, Syrnium whiteheadi, Scops fuliginosa, Hyloterpe whiteheadi, Siphia erythacus, Iole striaticeps, and Prionochilus johanne. Three of the new species are figured: Syrniwm white- headi, Siphia erythacus, and Prionochilus johanne. Sharpe, R. B.: Catalogue of the Passeriformes, or perching birds, in the collection of the British Museum. Fringilliformes: part IIT, containing the family Fringillide. Cat. Birds Brit. Mus. London 1888), 12, 1-872, pls. 1-16. Synonymy, keys, and descriptions of all the Fringillide known at date of publication. No Philippine species is figured. Sharpe, R. B.: On the ornithology of northern Borneo. With notes by John Whitehead. Jbis (1889-1890.) Seven parts with pages and plates as follows: Part I, (1889), 64-85, pls. 2-4; part II, (1889), 185-205, pls. 7 and 8; part III (1889), 265-283, pl. 9; part IV (1889), 409-443, pls. 11 and 13; part V (1890), 1-24; part VI (1890), 133-149, pl. 4; part VIL (1890), 273-292, pl. 8. This important contribution to the ornithology of Borneo contains numerous notes on species which range to the Philippines. Cryptolopha montis is figured in part II, pl. 8. Part VII, pp. 274-285, contains a table of distribution showing the relation of the Bornean ornis to that of the surrounding islands. 292 M’GREGOR. Sharpe, R. B.: Catalogue of the Passeriformes, or perching birds, in the collection of the British Museum. Sturniformes, contaiming the families Artamidexe, Sturnide, Ploceide, Alaudide, also the families Atrichiid and Menuride. Cat. Birds Brit. Mus. London (1890), 13, 1-702, pls. 1-15. Synonymy, keys, and descriptions of the swallow-shrikes, starlings, weaver- finches, and larks known at date of publication. No Philippine species is figured. Spodiopsar new name, page 665, to replace Poliopsar Sharpe, preoccupied. Munia cabanisi new name, page 353. Sharpe, R. B.: Coraciide, Meropide, Alcedinide. Cat. Birds Brit. Mus. London (1892), 17, 4-312 pls. 1-12. A review of the rollers, bee-birds, and kingfishers with keys, synonymy, and descriptions. Ceya steerti is described as new. The Philippine species figured are Hurystomus orientalis and Halcyon chloris. This is a very important work. Sharpe, R. B.: Rallide and Gruide. Cat. Birds Brit. Mus. London (1894), 28, 1-228; 248-277, pls. 1-9. The rails and cranes are described with keys and full synonymy. The work is an important one. Rallina ewryzonoides is the subject of figure 1 on plate 8. : Sharpe, R. B.: Catalogue of the Limicole in the collection of the British Museum. Cat. Birds Brit. Mus. London (1896), 24, 1-796, pls. 1-7. A very important review with descriptions, keys, and synonymy of all the Limicol. No plates of Philippine species. Sharpe, R. B.: Platalew (ibises and spoonbills) and Heroniones (herons and storks). Cat. Birds Brit. Mus. London (1898), 26, 1-328, pls. 1-5. An important review of these birds with descriptions, keys, and synonymy. The Philippine species figured are-Phoyx manillensis and Butorides spodio- gaster. Shelley, G. E.: Capitonide and Cuculide. Cat. Birds Brit. Mus. Lon- don (1891), 19, 13-121; 209-434, pls. 1-5; 11-13. The parts of this volume indicated above, dealing with the barbets andl euckoos, include keys, synonymy, and descriptions of all species known at the date of issue. _A useful and necessary work; no Philippine species is figured. AN IMPROVED METHOD OF MODELING ESPECIALLY ADAPTED FOR THE CENTRAL NERVOUS SYSTEM. (From the Anatomical Laboratory, Philippine Medical School, Manila, P. I.) PREPARATION OF BRAIN MODELS. By Marta Paz MenpozA AND MANUEL RAMIREZ. During the last summer session of the Philippine Medical School, Dr. Bean recommended to us the study of the nervous system and sug- gested that in the dissection of the brain we make at least three different sections, sagittal, coronal (frontal), and horizontal, and take an exact copy of the two surfaces of each section on pieces of blotting paper,’ cut out the ventricles and paste each two, which complete one section, together by means of small cubes of wood cut in such a way that the resulting thickness is just the same as that of the original section. The idea was to show the shape, size and relative position of the brain ventricles through the gaps left by the pieces of blotting paper pasted on the blocks, and on the blotting paper to show the internal structure of the brain. We adopted the suggestion and thought it capable of further improvement by substituting for the blocks of wood blotting-paper pulp, thus also gaining the external morphology and converting the work into true modeling. We found this to be an easy, economic, interesting and accurate work, the success of which is shown by the photographs accompanying this description. PROCESS. 1. Materials: Blotting paper, white and onion skin paper, gum arabic and paints. 2. Procedure: Preparation of sections. For coronal sections, an entire brain was taken from the preserving fluid,” held on a mass of cotton and ten sections of equal thickness made with a brain knife. Sagittal and horizontal sections—The brain is first divided into *Sussana Phelps Gage: The Method of Making Models from Sheets of Blotting Paper. The Anatomical Record. (1907), 7, 166. * Miiller’s fluid was employed so that the same material could be used for work on the histology of the brain. 293 294. MPNDOZA, RAMIREZ, AND ENRIQUEZ. halves, by making a cut through the longitudinal fissure, separating the brain-stem at the middle. One-half of the brain is then cut into four sagittal sections of the same thickness; the other half into four horizontal sections of the same depths. Thinner sections may be made, yet they are not desirable for beginners. Making the pulp.—Waste blotting paper that has been used is pulled apart by hand * into small pieces and put into water to macerate, care being taken first to wash the mass two or three times. When the pieces become soft they are reduced to a fine pulp with the fingers. The surplus water is then squeezed out and the pulp thoroughly mixed with gum which has previously been dissolved. Making copies of the contiguous surfaces of two adjacent sections.— The most anterior coronal section is first taken, held on a mass of cotton, the cut surface to be copied is then moistened with the preserving fluid and the onion skin paper immediately placed upon it. The struc- tures beneath the latter are now plainly visible and their outlines can be followed with a sharp pencil, thus making an exact copy of the posterior surface of this section of the brain and the anterior surface of the next. When this part of the work is finished, two pieces of white paper with two of blotting paper are taken, the onion skin paper placed over them and the external outline of the section cut through them. The blotting paper is then set aside and two pieces of carbon paper inserted between the pieces of white paper in such a manner that when redrawn, the outline of the section on the onion skin paper gives two copies on the white, representing the two adjacent surfaces, one of this and one of the next section. By proceeding in the same manner with the remainder of the sections copies of their surfaces can be obtained. Model preparation.—The work is now ready for modeling. The two pieces of blotting paper, set aside in the last preparation, which correspond to the anterior and posterior surfaces, for example of coronal section number 2, are taken, the surfaces to be put in contact with the pulp are painted with thick gum and between them such a quantity of pulp, previously squeezed of its surplus liquid, is placed so that its thickness is about 1 millimeter, for every centimeter, more than the original section of the brain. The less the water and the more homogeneous the pulp, the less will be the shrinkage of the model in drying. The border is now made even throughout with fingers, forceps, and probes; once all the sections are thus made they are put together in their natural position and the rough outline of the most important and deep fissures such as those of Sylvius and Rolando and the longitudinal fissure are marked. Afterwards, taking each sec- ’Not cut with a scissors or knife. AN IMPROVED METHOD OF MODELING. 295 tion individually, the sulci and gyri are worked out with the original section as a model, the two adjacent sections being occasionally consulted for their exact locations. Whenever the pulp dries too much it is moistened with a diluted solution of gum arabic. Painting the model.—To distinguish the white from the gray matter and the different structures inside the brain from each other, water colors are used for all but the cortex, where oil is preferable. While the model is drying the two outlined pieces of white paper corresponding to section No. 2 are taken and with the original sections at hand the different parts are painted with different colors, giving the same color to those parts having the same or allied structure. Last step.—When the model-sections are dried (which takes about one week, or longer, according to the water contained in the pulp), the white papers containing the copies of the surfaces of the sections are pasted on them and the gap left around the border, due to a slight contraction of the blotting paper, is filled wp with pulp and left to dry. Next, the cavity for the ventricles is excised, as this can not be done when the pulp is moist, the newly exposed parts painted and after all is completed the outside is painted with oil colors. The same procedure is followed for the coronal sections and the sagittal and horizontal ones. The record book.—A record book should be kept in which the onion skin papers containing the outlines of the internal structures should be preserved and to each of these the corresponding name should be given as their study is continued. Haperiments—With the hope of further improving brain modeling we conducted some experiments the results of which are as follows: Newspaper, macerated for three days in water, and reduced to a fine pulp of dark-gray color, has short fibers, but forms a soft mass easily workable into any desired shape. Painted with oil colors, almost no contraction occurs when drying and sections can easily be reduced to an 8 millimeter thickness. With one-half newspaper pulp and one-half lime a pulp is produced that can be used to make sections 8 millimeters thick; this dries within twenty-four hours into a hard, strong, white mass with an even surface, which is easily painted. With gum the resulting mass is more plastic, but dries less quickly. With one-third pulp, one-third lime and one-third sand, the mass is dark and dries more quickly, but it is brittle and not adapted for brain modeling. We repeated the above experiments with magazine paper and ordinary white paper and obtained almost the same results. This method of modeling can obviously be used for other purposes such as the reproduction of fruits, insects, etc., so that it can be made of advantage to botanists, zodlogists, paleontologists, etc. 296 MENDOZA, RAMIREZ, AND ENRIQUEZ. PREPARATION OF BRAIN MODELS. By Pio VALENCIA ENRIQUEZ. The blotting-paper pulp is prepared as described in the work of Mendoza and Ramirez. The two hemispheres of a brain are then separated with a sharp. thin-bladed knife, and four vertical-longitudinal (sagittal) sections of one hemisphere made, the cuts being 1.5, 3, and 4.5 centimeters, respectively, from the mesial surface. The mesial sec- tion is then placed on the table and an exact outline of the structures found on each of its two surfaces drawn through transparent paper. Two pieces of blotting paper are cut from these outlines. A sufficient quantity of the macerated blotting paper is then put on the table, and compressed with a flat board until it becomes somewhat solid and nearly as thin as the mesial brain section. Its exposed surface is now ready for a coat of library paste which is first put on thin, then thick, and the proper blotting paper outline is then attached to this by compression. The other surface is treated in like manner and the remaining sections are made in the same way. The exposed borders are then trimmed, the gyri, sulci, and ventricles being modeled while the material is soft and plastic, all exposed surfaces being coated with library paste. Drying—The models dry better in the shade than in the sunshine, rapid drying causing them to warp and to lose their shape. They should be placed in a closed locker over a pile of blotting papers on a plane surface. The model should be turned once or twice each 24 hours and the blotting papers should be replaced with dry ones at the same time, in order to facilitate the drying process. After the models are dry they may be painted in any way desired. Caution—All models made in this way must be preserved in closed cases, especially in the tropics, in order that rats, roaches, and other vermin may not mutilate or destroy them. Fie. Fic. Vig. 2 ae 9 ae ILLUSTRATIONS. PLATE I. . Mesial surface of model of right hemisphere showing separation of sec- tions. Mendoza and Ramirez. Superior view of the first section from the base of the hemisphere. Mendoza and Ramirez. Superior surface of the second section from the base of the hemisphere. Mendoza and Ramirez. Pirate I]. . Coronal section number 5, showing the posterior surface. Mendoza and Ramirez. Coronal section number 6, showing the posterior view. Mendoza and Ramirez. Model of the right hemisphere of a human brain viewed from the side. Valencia. Puate IIT. . Same as Plate JI, fig. 3, with the first lateral section of the cerebral hemisphere removed. Represents a sagittal section 4.5 centimeters from the mesial surface. Valencia. Same as Plate II, fig. 3, with the first and second sagittal sections re- moved. Represents a sagittal section 3 centimeters from the mesial surface. Valencia. With the first, second, and third lateral sagittal sections removed. Represents a sagittal section 1.5 centimeters from the mesial surface. Valencia. 297 MENDOZA AND RAMIREZ: METHOD OF MODELING. ] [PH1n. Journ. Scir., Vou. III, No. 4. MENDOZA AND RAMIREZ: METHOD OF MODELING. ] [PuHIL. Journ. Scr., Vou. III, No. 4. MrnpbDozA AND RAMIREZ: METHOD OF MODELING. ] [PHIn. Journ. Scr., Vou. III, No. 4. Fic. 2. Fic. 3. PLATE I1l. EDITORIAL. NOTES ON THE APPEARANCE OF SIREX JUVENCUS LINN. IN MANILA, P. 1. A few days ago a representative of a Manila importing house brought to this office an insect for determination which was found inside a packing case recently received from Germany. I have identified it as Sirex jguvencus Linn.* This wood-wasp is common in Germany and is well known to foresters as injurious to Pinus sylvestris Linn. The history of its introduction is as follows. A Filipino in opening a packing case noticed that about ten of the insects flew out, while at the bottom of the box several dead specimens were found. The packing “case contained cotton underwear, packed in pasteboard boxes, and it was noticed that the insects had eaten holes in these inner boxes and had slightly injured their contents. ; Unfortunately I did not see the packing case, but I have no doubt but that it was made of pine wood, in which the insects passed the last stages of their development. The wasps after hatching emerged on the inside of the box, and in attempting to escape attacked the contents of the box. W. SCHULTZE. NOTES ON THE ABUNDANT APPEARANCE OF GIBBIUM SCOTIAS FABR. IN THE PHILIPPINE ISLANDS. Some time ago Mr. Herbert 8. Walker, chemist of this Bureau, called my attention to the fact that there were numerous beetles present in a material known as argol.? Upon investigation I found enormous numbers of a beetle which I have identified as Gibbiwm scotias Fabr., together with some specimens of T’riboliwm ferruginewm Fabr. The *In German, “Die gemeine Kiefernholzwespe,” see Ratzeburg, Forst-Insecten (1844), 3, 148, taf. IV, fig. 3. * Argol (German, Weinstein) is a crude cream of tartar or potassium acid tar- trate which forms as a crust on the inside of vessels in which wine has been fer- mented. The color is purple to white, according to the kind of wine. It is used as a reducing agent in gold and silver assays. The purified cream of tartar is used in medicine and to a large extent in the manufacture of baking powders. Our supply came in 50-kilogram barrels from a firm in New York. 74196——_9 299 was infested with larve and cocoons of the first-mentioned species. As the barrel had been open for some time there was a possibility that the beetles might have entered it after it had reached Manila. I therefore examined a fresh barrel of argol which was kept in another part of the building. The outside of this package was in perfect condition and showed no signs of holes, but its contents were infested throughout. The surface of the argol was covered with a shiny mass of dead and living beetles of G. scotias Fabr. The above-mentioned facts convinced me that this species was introduced or imported with the argol from — the United States. This species probably found suitable climatic con- ditions and multiplied rapidly. In the literature which I have consulted I find no mention of this species having been detected in argol, although it seems to have a somewhat peculiar sense of taste. Some authors mention it as feeding on the dust of granaries;* Laboulbéne * found it on an Egyptian mummy in large numbers; Lucas ® saw larve and adults feeding on red pepper (Capsicum annuum Linn.), and Stierlin ° on the wool of sheep. Should this beetle at any time adapt itself to food of more economic value and importance than the variety mentioned, serious damage might — arise. W. SCHULTZE. ® Boieldieu, “Monographie des Ptiniores,” Ann. Soc. Hint. de France (1856) (3), 4, 679. f 2 * Loc. cit. (1872) (5), 2, 29 (Bull.). 5 Loc. cit. (1884) (6), 4, 76 and 124 (Bull.). ®°Calwer’s Kdferbuch (1893), 397. 50-kilogram barrel of argol was nearly full and throughout the mass it : Vou. III NOVEMBER, 1908 No. 5 THE PHILIPPINE JOURNAL OF SCIENCE EDITED BY PAUL C. FREER, M. D., PH. D. WITH THE COOPERATION OF MERTON L. MILLER, PH. D.; GHORGE F. RICHMOND, M. S. SOW Da SMITHS Pa Dr Aa he COX, PHD: RAYMOND F. BACON, Px. D.; CHARLES S. BANKS, M. S. H. D. GIBBS. B. S.; R. C. McGREGOR, A. B. PUBLISHED BY THE BUREAU OF SCIENCE OF THE GOVERNMENT OF THE PHILIPPINE ISLANDS A. GENERAL SCIENCE MANILA BUREAU OF PRINTING PREVIOUS PUBLICATIONS OF THE BUREAU OF GOVERNMENT LABORATORIES, 1No. 1, 1902, Biological Laboratory.—Preliminary Report of the Appearance in the Philippine Islands of a Disease Clinically Resembling Glanders. By R. P. Strong, M. D. No, 2, 1902, Chemical Laboratory——The Preparation of Benzoyl-Acetyl Peroxide and Its Use as an Intestinal Antiseptic in Cholera and Dysentery. Preliminary Notes. By Paul C. Freer, M. D., Ph. D. 1No. 3, 1903, Biological Laboratory.—A Preliminary Report on Trypanosomiasis of Horses in the Philippine Islands. By W. HE. Musgrave, M. D., and Norman E. Williamson.’ 1No. 4, 1903, Serum Laboratory.—Preliminary Report on the Study of Rinderpest of Cattle and Garabaos in the Philippine Islands. By James W. Jobling, M. D. 1No. 5, 19038, Biological Laboratory.—Trypanosoma and Trypanosomiasis, with Special migtorence to Surra in the Philippine Islands. By W. HE. Musgave, M. D., and Moses . Clegg. 1 No. 6, 1903.—New and Noteworthy Plants, I. The American Element in the Philip- pine Flora. By Elmer D. Merrill, Botanist. (Issued January 20, 1904.) 1No. 7, 1903, Chemical Laboratory.—The Gutta Percha and Rubber of the Philippine Islands. By Penoyer L. Sherman, jr., Ph. D. 1No. 8, 1903.—A Dictionary of the Plant Names of the ESRD Islands. By Elmer D. Merrill, Botanist. 1No. 9, 1903, Biological and Serum Laboratories.—A Report on Hemorrhagic Septi- cena na ‘Animals in the Philippine Islands. By Paul G. Woolley, M. D., and J. W. Jobling, M. D. 1 No. 10, 1908, Biological Laboratory.—Two Cases of a Peculiar Form of Hand Infection (Due to an Organism Resembling the Koch-Weeks Bacillus). By John R. McDill, M. D., and Wm. B. Wherry, M. D. 1No. 11, 1903, Biological Laboratory.—Entomological Division, Bulletin No. 1: -Prelimi- nary Bulletin on "Tnsects of the Cacao. (Prepared Especially for the Benefit of Farmers.) By Charles 8. Banks, Entomologist. 1 No. 12, 1903, Biological Laboratory.—Report on Some Pulmonary Lesions Produced by the Bacillus of Hemorrhagic Septicemia of Carabaos. By Paul G. Woolley, M. D. No. 13, 1904, Biological Laboratory.—A Fatal Infection by a Hitherto Undescribed Chromogenic Bacterium: Bacillus Aureus Fotidus. By Maximilian Herzog, M. D. 1 No. 14, 1904.—Serum Laboratory: Texas Fever in the Philippine Islands and the Far East. By J. W. Jobling, M. D., and Paul G. Woolley, M. D. Bvtological Laboratory: Entomological Division, Bulletin No. 2: The Australian Tick (Boophilus Australis Fuller) in the Philippine Islands. By Charles §. Banks, Entomologist. No. 15, 1904, Biological and Serum Laboratories.—Report on Bacillus Violaceus Ma- nile: A Pathogenic Micro-Organism. By Paul G. Woolley, M. D. 1No. 16, 1904, Biological Laboratory.—Protective Inoculation Against Asiatic Cholera: An Experimental Study. By Richard P. Strong, M. D. No. 17, 1904.—New or Noteworthy Philippine Plants, II. By Elmer D. Merrill, Botanist. 1No. 18, 1904, Biological Laboratory.—I. Amebas: Their Cultivation and Etiologic Significance. By W. E. Musgrave, M. D., and Moses T. Clegg. II. The Treatment of Intestinal Amebiasis (Amcebic Dysentery) in the Tropics. By W. E. Musgrave, M. D. Now19, 1904, eee aay Laboratory.—Some Observations on the Biology of the Cholera Spirillum. By W. Wherry, M. D. No, 20, 1904. W: Brolvgteal Laboratory: I. Does Latent or Dormant Plague Exist Where the Disease is Endemic? By Maximilian Herzog, M. D., and Charles B. Hare. Serum Laboratory: II. Broncho-Pneumonia of Cattle: Its Association with B. Bovisepticus. By Paul G. Woolley, M. D., and Walter Sorrell, D. V. S. III. Pinto (Pao Blanco). By Paul G. Woolley, M. D. Chemical Laboratory: IV. Notes on Analysis of the Water from the Manila Water Supply. By Charles L. Bliss, M. S.. Serwm Laboratory: V. Frambeesia: Its Occurrence in Natives in the Philippine Islands. By Paul G. Woolley, M. D. No. 21, 1904, Biological Laboratory.—Some Questions Relating to the Virulence of Pe Sar -Organisms with Particular Reference to Their Immunizing Powers. By Richard Strong, M. D. = No. 22, 1904, Bureau of Government Laboratories.—I. A Description of the New Build- ings of the Bureau of Government Laboratories. By Paul C. Freer, M. D., Ph, D. If, A cotalbene of the Library of the Bureau of Government Laboratories. By Mary Polk, ibrarian. ' - 1No. 23, 1904, Biological Laboratory.—Plague: Bacteriology, Morbid Anatomy, and Ebstop aha (Including a Consideration of Insects as Plague Carriers). By Maximilian erzog, M. D. No. 24, 1904, Biological Laboratory.—Glanders: Its Diagnosis and Prevention (Together with a Report on Two Cases of Human Glanders Occurring in Manila and Some Notes on the | Bacteriology and Polymorphism of Bacterium Mallei). By William B. Wherry, M. D. No. 25, 1904.2—Birds from the Islands of Romblon, Sibuyan, and Cresta de Gallo. By Richard C. McGregor. ; yiky No. 26, 1904, Biological Laboratory.—The Clinical and Pathological Significance of Balantidium Coli. By Richard P. Strong, M. D. No. 27, 1904.—A Review of the Identification of the Species Described in Blanco’s Flora de Filipinas. By Elmer D. Merrill, Botanist. No. 28, 1904.—I. The Polypodiacer of the Philippine Islands. II. Edible Philippine Fungi. By Edwin B. Copeland, Ph. D. No. 29, 1904.—I. New or Notewsrthy Philippine Plants, III. II. The Source of Manila Elemi. By Elmer D. Merrill, Botanist. No. 80, 1905, Chemical ‘Laboratory. —I. Autocalytic Decomposition of Silver Oxide. II. Hydration in Solution. By Gilbert N. Lewis, Ph. D. No. 81, 1905, Biological Laboratory.—I. Notes on a Case of Hematochyluria (Together with Some Observ ations on the Morphology of the Embryo Nematode, Filaria Nocturna). By William B. Wherry, M. D., and John R. McDill, M. D., Manila, P. I. II. A Search Into the Nitrate and Nitrite Content of Witte’s ‘‘Peptone,”’ with Special Reference to Its pnriuericeson. nae Demonstration of the Indol and Cholera-Red Reactions. By William B. erry, M. D. 1 Out of print. 4 The first four bulletins in the ornithological series were published by the Ethnological Survey under the title “Bulletins. of the Philippine Museum.” Later ornithological publications of the Government appeared as publications of the Bureau of Government Laboratories. (Concluded on third page of cover.) THE PHILIPPINE JOURNAL OF SCIENCE A. GENERAL SCIENCE Vou. TIT NOVEMBER, 1908 No. 5 PHILIPPINE COALS AS FUEL. By Atvin J. Cox. (From the Laboratory of Inorganic and Physical Chemistry, Bureaw of Science, i Manila, P. 1.) INTRODUCTION. While it may be true that the testing of fuels under boilers at best gives only approximately comparative results, nevertheless there is no degree of accuracy in assumptions such as that of Bazin,’ who considered the practical steam-making capacity of a combustible material to be two-thirds of its found heating value. This capacity may vary from 80 per cent with the best anthracite down to 50 per cent or eyen less when a highly bituminous coal is used. The type of plant, the personnel and other important factors must be considered. The error in concluding that a coal high in eyaporative power is on that account the best coal and conversely that a very cheap fuel necessarily must be cheap in the long run should be guarded against. The most satisfactory way in which a correct conclusion as to the respective commercial values of different coals can be arrived at is to make tests and then compare their performances as shown below. There is no doubt that steam vessels can successfully use some of the Philippine coals. If others are too high in volatile combustible matter _ they unquestionably can be employed by mixing them with a certain amount of Australian coal and thus too rapid gasification be prevented. The Coast Guard and interisland ships now burn on the average 10 tons of Australian coal each per day or 300 tons per month. If they replace *Rev. gen. de Chim. (1904), 7, 91; Rev. in J. Am. Chem. soc. (1905), BQihy Uses}, 75034 301 302 COX. all or two-thirds by Philippine coal it would require only a simple calculation from the following data when the prices per ton are known, to determine the difference in cost. I know of but one trial of the commercial value of Philippine coal where complete data of the test were kept. This was made about two years ago at the Philippine Cold Storage and Ice Plant.* The test was as satisfactory as possible under the existing conditions; the results exceeded the anticipations of those in charge of the test and seemed to indicate “its equality with many other coals on the Manila market.” However, the grates were not adapted to the fuel and much incon- venience was experienced because the decrepitated coal passed through the ~ grate with the ash. Toward the end of the test, this ash was burned over again and after the second burning the analyses of this Bureau showed it to contain 62.6 per cent of combustible matter. No com- parative tests were made with other coals. In 1904 the United States Army transports Chukong, Sacramento and Palawan made runs on Batan coal and the reports in each case were favorable. The coal was easily fired, it burned well, the amount of soot was comparatively small, there was no great quantity of smoke, the content of ash was low and there was no clinker. The object of the following investigation was to determine the steam- making value of the coals of the Philippine Islands, as measured by kilos of water evaporated per kilo of fuel when used under a boiler, as compared with the foreign coals offered on the market in this Archipelago ; it has also been my purpose to make a comparative study of the individual coals as well as to convert into useful work the greatest possible per- centage of heat units contained in each. Careful and complete records have been preserved of each test; therefore it should be possible for engineers to determine from the data which are given whether or not the conditions were those best suited to the coal under examination and when a price is established for these coals, these tables will form a basis of comparison not only as to the water evaporated per kilo of fuel, but also in regard to the water evaporated per peso of fuel cost. In com- mercial operations the all important question is to find the fuel which will run a plant with the least financial outlay. A special grate was tried for some of the coals and an effort has been made to use a method of firing which would give the best results. As the supply of material at my disposal was limited, except in the case of Australian coal, only a small amount of preliminary experimenting could be done to determine the best practice in regard to firing and to gain information regarding the fuel before beginning the test. An engineer always needs experience with a coal to burn it in the most efficient manner. It will be noticed from the tables that in some cases *The Far Eastern Review, January (1906). PHILIPPINE COALS AS FUEL. 303 the efficiency for the second run is slightly higher than that for the first, showing the benefit of the first day’s experience; however, in no case is the difference much greater than the possible error from other sources. Several preliminary trials were made on the coal regularly used here for firing in order thoroughly to test the working condition of the apparatus. It would have been very desirable to have had duplicate determinations of the steaming quality of each coal, but this was not always possible with the supply on hand; nevertheless it is believed that all the results are complete and sufficiently reliable to show the nature and indicate the real fuel value of the coal; in fact it has recently been shown * that more than one test of a coal is superfluous. Seventy- seven first tests gave an average efficiency of 66.05 and seventy-seven second tests an average of 66.02 and thirty-two third tests one of 65.87. Tt is evident that promiscuous tests made under different conditions are not at all comparable, for it would be impossible to discover whether the variation was due to the fuel, the apparatus or the manipulation. However, in the work done at this Bureau many factors have been eliminated by using the same plant * and the same personnel; the others have been carefully controlled by using the same apparatus and main- taining all manipulations and general conditions as nearly uniformly constant as possible, except where a change in the second test was to the advantage of the. coal. With the variable factors eliminated, the coals can be directly compared. DESCRIPTION OF APPARATUS AND METHODS EMPLOYED. All instruments used were carefully standardized and every precau- tion taken to prevent the possibility of error. As the nature of the coals to be burned was so entirely different, two sets of grates were provided. The one was of plain, single bars 1.5 centimeters in width and constructed to give an air space of 1.2 centimeters between each pair, or a ratio between air space and grate surface of 20 to 45. The other, constructed for these tests and used with some of the coals, was a perforated grate with round, tapering holes 1.25 centimeters in diameter at the top, the smallest dimension, averaging 25 per square decimeter and giving a ratio between air space and grate surface of 18 to 45. The two boilers shown in Plate I are exactly alike, the following description applies to both; however, with one exception, the tests were made with the one on the right; they can afford only a clue as to the efficiency of the boilers. This was not sought, for there are no means of comparing the boilers with others fired with Philippine coal, or perhaps with themselves under different conditions. The boiler was thoroughly * Breckenridge, L. P., U. S. G. S. Bull. (1907), 325, 32. ‘The losses through radiation and conduction do not vary greatly for any given installation. : 304 COX. cleaned before beginning the test; it was in all cases used on the previous day so that the brickwork was thoroughly heated, and it was under full steam for some time on the day of the test before beginning the actual run. The gauge glass of each boiler was graduated into millimeters and calibrated independently with water at 30° C. These data were used to correct the water level between starting and stopping rather than by use of. the pump. BOILER: Kind, Babcock and Wilcox. Nominal rating, 75 horsepower. Type, water tube. Tubes: Number, 45. external, 10.16 centimeters. internal, 9.48 centimeters. Length exposed, 42.67 decimeters. Drum: Diameter, external, 9.15 decimeters. Length, external, 58.4 decimeters. Diameter i e > iS) Water-heating surface— ecumere OPP ese eee EE ERIS OS ye CRs VRE ered Cee 5,715.2 OP drm eee ee ER ea ad eee 748.8 ED tallies Pi S Sieg Ne Ya ED Rear een Re ae te a ea 6,464.0 Steam gauge, Ashcroft’s, graduated to 5 pounds on a 12-inch dial. FURNACE: Kind, Hand fired. 3 front, 12.2 decimeters. AM (eae 8.3 decimeters. Width, 9.90 decimeters. Flue connecting to chimney: Length, 18.3 decimeters. Calorimeter, 49.4 square decimeters. Grate: Kind, gridiron bar or perforated as best adapted to the in- dividual coal. Width, 9.90 decimeters. Length, 18.3 decimeters. Area, 181.2 square decimeters. Ratio of water heating surface to grate surface, 35.7: 1. CHIMNEY: Diameter, internal, 12.2 decimeters (4 English feet). Height above grate, 30.5 meters (100 English feet) . Area, 38.33 square decimeters. The stack was high enough in all cases to give the draft necessary for the coal in the condition used. Draft, natural. We have no economizer. The exhaust main passes through a 200-horsepower Wainwright even-flow feed-water heater. PHILIPPINE COALS AS FUEL. 305 During all of these tests the steam was used to operate a large duplex steam pump, to drive the engine which furnishes the power to operate the air compressor, the vacuum pump, the refrigerating machine and many small motors, ete., for the laboratory and to supply live steam throughout the building. At first I imtended to take switch-board readings, but the idea was given up as impracticable. Owing to the intermittent use of steam for other purposes such readings would neces- sarily be incomplete; but in Plates II to VII, I have given photographs of the volt meter and ammeter indicator diagrams. An estimation from these shows that an average of about 60 per cent of the steam produced was used by the engine, and 40-per cent for other purposes, including that condensed by radiation from the pipes. The equivalent evaporation per indicated horsepower was assumed as 25 kilos of water, because of the ight and variable load of the engine. The portable drop-lever Howe scales used in making the weighings were care- fully standardized and found to be correct; the meter was fitted with a gauge and regulators so that it was calibrated from time to time by actually weighing the water passing through under the same head as it was fed into the boiler and no error was at any time detected in the registrations of the meter. If there was a slight error, being constant, it would affect alike all the tests and therefore be negligible in securing data for comparative purposes. The boiler feed pump was run intermittently and always at the same rate. The tem- perature of the water entering the boiler from the heater was determined by readings of a thermometer placed in a thermomettr cup on the pipe just adjacent to the boiler. The steam gauges were tested by comparing with the test-gauge of the Crosby Steam Gauge and Valve Company, a standard instrument manu- factured by Schiiffer & Budenberg, Limited, and that used by the city boiler inspector. The only errors were in the initial setting of the needles. These in all cases were corrected at a pressure of 20 pounds per square inch by actual trial with a column of mercury. The damper was controlled by a lever passing over a graduated segment. The chemical thermometers were of 550° C. capacity, and were calibrated by the Physikalisch-Technische Reichsanstalt in Charlottenburg, Germany. The tem- peratures of the flue gases were read from a high-grade mercury thermometer which was calibrated from these. The usual U tube, or inverted siphon of water, draft-gauge was used. One arm was open to the atmosphere and the other, by means of the proper connections was inserted into the draft to be tested. The difficulties of reading the gauge were reduced to a minimum by the looking-glass scale. The latter was accurately divided into millimeters so that the error of reading was not greater than a few units in the decimal. The scale was movable, which greatly facilitated the reading of it. A Barrus’ continuous, surface condenser calorimeter was on hand during several of the tests to determine the moisture in the steam. Steam nearly always carries water with it and thus the boiler is credited with having evap- orated more water than is really the case. However, the results recorded in Table II have not been corrected for this since I was unable to determine the factor for all. It will be seen from the following table that the boiler of this Bureau produces steam which is very uniform in quality and as the results 306 COX. of the tests are intended to be comparable only, it is permissible to omit this constant correction entirely. It was not convenient to attach the calorimeter close to the boiler. It was attached to the steam pipe 22 feet away and owing to the radiation from this pipe, even though all parts were well covered, the amount of moisture may be somewhat high. The readings were made on several days during the firing of coal from three different sources and at different times of day, so that the greatest variations of load are represented. These readings are shown in Table I. TasLe I.—Stcam calorimeter readings. Readings of Time Steam- thermometer. Date, 1907. after gauge starting. | pressure. Upper. | Lower. Kilos per h. m.| sq.cm. cC: CG: June Hoes 5 20 7. 8044 168 109 5 45 7. 8044 168 110 6 10 7. 8044 169 110 6 35 7.1716 165 -| 110 June!20 sees sseeeene 1 12 7. 8044 167 109 1 50 7.5232 167 109 2 42 6. 8200 164 110 7 3 00 7. 8044 168 110 4 00 7. 8044 168 110 4 40 7. 8044 167 109 4 55 6. 4685 161 108 6 25 7.3825 166 1 JUNE 21 eae arene 2 40 7.5232 167 110 3 20 7.1716 165.5 | 109.5 5 50 7.5232 167.5 | 108.5 6 20 7. 5232 167 108.5 6 30 7.1716 166 108 6 45 7. 8747 169 109 Julyplbyesseseesasenae 5 30 7.1716 163 110 5 50 7.1716 163 110 July 1 30 7.1716 164 110 1 50 7.4529 165 109 4 40 7.3825 165 110 5 05 7.1013 163 109 5 20 7. 8747 167 110.5 6 15 7.5935 166 110 PHILIPPINE COALS AS FUEL. ; 307 The average readings on all of these tests are, for the steam-gauge pressure 7.4501 kilograms per square centimeter (106 pounds per square inch), for the upper thermometer 166° (330°.8 F.), and for the lower thermometer 109°.4 (229° F.); the ranges for the thermometers being from 161° to 169° and from 108° to 110°.5, respectively. The normal for 166° may be taken as 139°.5 and the average cooling below this point is 80°.1. In order to compute the amount of moisture from the loss of heat shown, the number of degrees of cooling is divided by the coéfficient, which depends upon the specific heat of steam, representing the number of degrees of cooling due to 1 per cent of moisture. 30°.1 divided by the coéfficient given by Barrus® for 166°, which is 11°.66 (21° F.), gives 2.58 per cent. No correction has been made for the moisture produced by radiation from the apparatus itself. Denton ° has shown that it is seldom possible to operate a condensing calorimeter with the degree of exactness calculated for the instrumental error, namely, 1 per cent. There has always been found to exist ac- cidental variation considerably in excess of the theoretical instrumental error, even Regnault’s experiments, the results of which are presented in tabular form in most publications upon the properties of steam, being no exception in this respect. He has also shown that jets of steam show unmistakable change of appearance to the eye when steam varies less than 1 per cent, either in the direction of wetness or of superheating, from the condition of saturation. When a jet of ‘steam flows from a boiler into the atmosphere under such conditions that very little loss of heat occurs through radiation, etc., the quantity of water if not too large, may be estimated from the color of the steam. If the jet is transparent close to the orifice, or is even a grayish-white color, the steam may be assumed to be so nearly dry that no portable condensing calorimeter will be capable of measuring the amount of moisture in it. If the jet is strongly white, the quantity of water, if it does not exceed 2 per cent, may roughly be estimated, but an amount in excess of this can be determined only with a calorimeter. Careful observations roughly corroborated the results given in Table I. Analysis of the flwe-gases—The flue-gases were analyzed and for this purpose were drawn from the flue by means of a sampler such as is shown in the following figure: ° Trans. Am. Soc. M. E. (1890), 11, 795. °TIbid (1899), 10, 326. 308 COX. Flue wall -->77- OD arunes COG OUND |S O Diss | eae oan ak ete ere APES Sas IGT Carelag tL 9 °S6L 0698 6 ‘80T 9¢9°L 69°66 | 02°FQL | Lo | Or SOG TRACES Nn | iennnnnnnnnn Chara, [PSS p ‘ON Wes ‘UO BAIOSEL AIB}ITTN &I T Str 6L°8 FOIL T9L°L 06°62 | Gh6GL | Lo | OD asa ciate wu? | (1) Opsieney | grea Caer aE AEE EN TUG cage eae od rag T Ser c6L°8 FOIL T9L°L (WAG | API, Ra OD mem SO6T ‘IS “IB | O Dina erie | per cred eee meen SF he ee Cae PE od IL @0¢L Toh 8 ¢ ‘cor LIP*L 96°6% | 48°09L ); th 9 | OPsss 5] SO6L FL LUBE | ~HOVIS POR AUNT |S a oUOlBAIOSOI AIBA or :puBISsS]T UBB P'6IT 668 '8 LF0T cog 2 QA | GES ND OQ 7,06, 9 eC Cn | ene OD cea nal cane Ge Eee TRS Teoh iit a ee Q)r | 6 6 TIT 898 °L C16 FE8 9 O886G) 9) F659S2)|| 2) tienen: ODingsias LO6L Gt Atng |----—~ CUM TOMGOd Dice cacinnaan aaa erm aa uBngey ‘ooulog | 8 6 PSE F8L'8 @ OIL OGL *L $9°6Z | 06°7S4 | G 9 5 (OD aaaies 8060 ‘82 ABI [>> C0) a ee ere aa (QouTAOIg OpIeyYOH) weqnz L T ‘921 C98 °8 PIL TE8°L $866 | 80892) 2 lisesi OQ] 806 ‘GS ‘IdW | dum |-----~ pups] nysnry ‘(ns}e1vy) tUBJOUTYSOR 9 :uede Lara ¢09 8 L°LoL TLG*L 98°6G | 8P°8SL | #9 9 | op 806E'2 ABE | ODEs Been Ou CR ahs ocean ees ee ae a OM ‘ ¢ ZLZL | OF6'S GOIL 906 °L GI | GBI OD aaa 80619 ABI |-~HOVB[s puw dum] | ous baa AOTIBA MOzyorT i 9 °S¢T T&8 ‘8 6 OIL L6L°L O8"62 | F6s9GL | 2 —|inne pay OD smaaa g061 ‘¢ oun |-~~~dumny] payoayag | eR ees cod € 8 °LIT E86 °8 T80L 6FGL B66 | 68°LGL | Lo | UOIIPID | LO6L ‘Te oun |------ OD rrgis Ratan rae REC ta ran ios Oh ihe aes mea g ZG Z'OZI_ | OGF'8 ¢'SOT OIF 'L €8'62 | QL°LGL | 2 ~~ pozB10}1od | L061 ‘0 oun |-~HoB[s pus dunyT fPUASPTBA\}SOM T ‘spunog | ‘sojiy |"spunog | ‘so2.y “SunoeT PBT BIISNY our Saas ue eta “s19}0UL z ol arsnbs | “Wout | orbs |'S?4°81) Airy : Tod orenbs lod ; “pasn : “1809 Jo : geet Tod : A ayer3 jo pury | PHt 30 8’ | oz1s peroromM09 Teo oveoinos ie —saj}nposqs —osnes “SuIpBal -BInd ainssoid &q oinssoid Jayomoisq AInd WBd}S OSBIOAY W1B9}S JSBIOAY -I0U O3BIOAY [‘sdsoulsuq [eoTURyOSW JO AJoID0g UeoTeIY 24} JO S1aquINU epod a1e SUUIN{OD IAAO seINnSy peovj-HoV[G 94.1] Jays DjWDdIY ay} UO ayvs sof pasafJo siayjo puv syvoo auiddywyg fo sjsaz buruva;9— |] WIV], COX, 312 ‘8Té ‘LTE ‘dd ‘a[qe} oq Jo pus oy} YV MOT[OJ Sajouj00,T Ral ea TS CAE Same 00's eee eaneeslie ema TGS £'88 63 ZB9Gx 6 Gz, Bay tind anette LOG (SP I er eine ee UMOIg | 08'S | PPP OF 68 98 '%G | 89% P09 ze L'98T 9G es oN eaama ea 1% 61 SOC Sign lPonspauskea ea TODS ss F859 || F656 G6 "18 1o'9P | GPS S'S Ke G SLL 0°08 se ltaass =a 6 8 PGGy[ye. 3 BLO AN) Senn eenn a Sanaa CD imme 94°¢ | TO‘OL £6 °LE 0g'9F | 068 1°08 Ke GCL SGC Hevea PISA 6 he SP. Re Op---~-| g0"0T | 19ST 0G 98 98°F | OPP ‘S°9L 8% 9 TLL £18 hea era eeca | 9 ee UMOIG YSIppoy | 80°L | £0'ST 9¢ 98 ee'se | 90F 6°LL 66 L Pel 408 T TOL ce LST aSieasecaig rine Ope = 1g9°% | 909 08 68 sete | 268 ZGL 63 0'PLT OOS Hach ines 6 val GSS on meena: Avis YSIppay | 70's 80°9 GL '68 CLT PIF 9'OL 6% 6 CLL Lard 6 8 OSiL.||ite sar conan eames OD saama #8'F | 489 66°88 og'0¢ | Es 6°8L 62 G'SLT BOSS i |Gaamamey 6 aL LS SSS UMOIG YSIPpoy | Ch'9 | 88"°G 9% 88 Ip'6b | O18 6'8L 66 G ‘SLT Chl Gases |aaaees IL IL OS TOSBIOAB ‘OTCBIIBA |~--7-- ABIS YSIPpAY | GG'g 8I'¢ 9L OP Tg "Gh PLP P'6L Ke 8 ‘TAT (GAG we \),| lear eeian eases oer a OL (GG lS oem eeredes eee OD ania 19°% | Sh'S og TF co0¢g | #868 8'9L 8% @'TLT 6°82 T IL 6 GGaIs| Pisces Sas pel qusrT | 19's | Shs cg TP gcog | 118 G°9L 8% 8891 OMTES Sisal aeaaesial IL 8 “LET OSBIOAB ‘OTQBINBA | UN | 68'0L | 281 09 GP 69'3F | G68 0°0L 62 P'SLT OVATE |e aera scenes L "BST OSVIIA ‘BTQBIEBA [TT OQ Tene len segue 8G°LE S8'8h | Ses 8'bL 8% 8 ELT ph Che. Ol oes | arenes sare 9 = temas. ie cisweae Op---~"| 08 er | TL% LPS 29'7G | 00F els 08 9°CLL 808 Peres sat [len G gece ay aT Op--—-| 082 | IL% LYSE z9'2¢ | Té6e lard 0g ‘PLT 018 Renmaaea| "eS: i Ledge unpre er a Avy | 606 | PLT #998 Speco | sue PG 62 9°SLL G6 T 6 & ROX G (ipsa armrest eR (ODEanes £0°ZT | 08% 8% FS #6'0G | £98 £08 8% 0'TLT Od NESS IL z OFT OSRIOAR ‘OTQBIIVA | AVIS YSE{Ud | gO'ZL | 08% & FE F60S | 098 Z'6L 8% 8 TLL S26 Cyan leases aL I Do Do Do Do Do cs aE 1% 0% 8st LT ST el “10}7 BUL ae, "10JvoU, 9 ¢ . ‘ysy foaysrone] UUHSNA |wonsw] Thy HECHT 3AU104) srtoquo | noted lrg Mer] yl | sede | ysoy Ree “ysv Jo 10[09 aTABIOA Surdeosgr Toye poo 1OJBM paddy) ‘WIRES uddeM eg 2 *19] BAM JO S19}9UL q' [B09 9] JO SISA|BUB 9] BUIIXOIg —Jo oinjyeroduie} o8B10AY -T[ [LU UT WJBIp U JO 90105 OSBIOAY Cite ‘d oes ‘[ets} JO uoTzVINp puv ‘pasn 9}VIB JO PUL ‘[Vl4} JO ayVp ‘[vOd Jo OZIS [VIDIBWIMIOD pUL BdINOS 10,7 J “‘panuryuog—jay.nw vpunyy ay7 Wo apps sof pasaflo svayjzo pun synoo amddyryg_ fo sjsa, burwya,39— TJ ATAV I, 313 PHILIPPINE COALS AS FUEL. ‘STs ‘LTE ‘dd ‘orqe} og} Jo pus oy} 42 MOTLOJ $9}00}00,T FES me Pere ED Oe SE BE | CSG LY jsxparinr 1 ea reanee ite. seer i| Pas oe Rees |e Sug oopcis nal ment ree | eae arary | eee ees || eames gi 0G TOROOn =| $9°La € GPL $988 ‘LT | 6°6LL F'LIL ‘ST | 9ce ‘2 690 ‘2 GL '9 LL8‘9 08 °S 19°06 &8'T 18°F OL 69 61 1620 “0 F0'T P S81. €OLP ‘LT | 8°80 0°99G°T | 90% ‘L 682 ‘9 690 °9 B69 8G 19°C SPL 68°¢ F889 8 6080 0 OL'T 8 002 GFOL‘L | 7802 PL2G TL | 800'L 9FL ‘9 TL0°9 L819 91°G 61°26 eh 06°¢ ¢s"69 LT. 9TFO'O Sh 0°69, 6'788'T | G"0se P'STs'% | 0689 09 ‘¢ 09g “F S19 ‘F £0 ‘OL TAG SIH 160 60'S 90 TS pe 66800 82 T FTEs 4°66 G88 O'FSI'T | 60z‘9 699‘G 169 F L19‘F 80°L #80 GL Se £60 98 °¢ 02S cL 6620 °0 10°T. T 861 S618 T | ¢ "coz SPORT. | 80° G80 ‘L 0199 pag ‘9 19% 610 60° LUT gh'¢ 6G 69 PL 9280 °0 9L'T 0'TIZ GOLF ‘LT | 9'F2a GCLG'T | eba‘L O10‘ #86 ‘9 ggg ‘9 FLO 01% 9L'T IF¢ £69 &L 6780 0 9L'T T'01z 6 OLF‘T | Z€2e 9°29S'T | Gar‘L eg ‘9 668 ‘9 696 ‘9 60°0 PS ‘0G SLT 29'F 84°89 ral 9280 '0 9L'T 6 01Z Z'POR TL | Lhe g8e'T | PST ‘L 169 9 862 ‘9 909 GP '9 60°0 8L0Z IL PoP €9°19 IL T8800 98 'T ZOPG 6 OFS T SLT | 9F0‘L TIP ‘9 610°9 92 ‘9 cG's OL‘ 86 6T 96°0 TPG 00"e9 Or £6800 OPE 0 FSG L°869'T | 0°69% 0164 'T 102 $99 ‘9 10¢ ‘9 19°% P40 SP 1G IL'T Gh'G 6G "99 6 £4800 isc 0'GES 1°869'T | 6°Gcz O16 'T L¥0‘L #99 ‘9 10¢ ‘9 19°% PLO Sh 1G Itt Gh'¢ 6G "89 8 12600 Lead L1G L°8G0'T | 9°FT% 6320 ‘T COG ‘L LOL‘L 198 ‘9 66 OT IL'0 &L IT 6L'T 02°F 86 GL L PS80°0 90° T 626 L'FO9‘T | bess 0'P89'T 918 ‘9 169 ‘9 ie) 18 ‘2L 160 TP 8 20°T ary 00°29 9 6620 °0 90°T G6 $908 ‘I | 2°260 G-pss T Set‘ L860 ae | aaa aR 08 “Gr ISS Putco opm ag oi ene ice Preece or G 9620°0 cO'T 9° T6T 9°GFL‘T | 2°S6T P PAL T Ser ‘L LSGLO era iea=cgt na 08 “GL BQN ON Ne eceeeicer sce [ices yer mal Came ane | ees See 2 89200 96 °0 @ SLT €°31Z T | § GAT 8°88 'T LOT‘L £86 ‘9 810 °L 616 Sr‘0 98 ‘IL 19°T. 18° 98 "SL & TZ80'0 PLT 6 °L06 T'Gch TL | 6 "81% 0°L6F ‘T 08 ‘9 P19 ‘9 GLE ‘9 £0 °ZE 60°0 ards SPT 19en2 9T OL z LIg0°0 SLT L P06 LCkh ‘TL | 9°01% O°PLP'T cos ‘9 PL9‘9 CLG ‘9 £0°CL 60°0 Sarde SPT IL'P OTOL T “SO7LY “SOL "SO]UM “SOM *SO]UM “SOU *$92.L0]D) | “89240709 | *$91.L07D9 “moy aod SP oP LG Fc} Tg og 98 1O]TBMA 90ByINS [BIOL -snq aecat ayy ULOIy anydtng HIRES) “Moy : susto}o) paye[noywo 480} 10d cere 3 r ‘poly SB = ea Jo oap avenby a MSoHIO [RO ON ais g “pory SB [BOD reba So EA ROO meats “[wod aT} JO SISATRUB OYVUITA *POUINSUOD o [BOD AIC, [Ite -d 00s ‘[et4} Jo UOTeVINp pue ‘pasn a}e1S JO PULYy ‘[ela] JO yep ‘[voOd Jo azIs [eIOLOMIMIOD puv 9dIN0s 10,7] “ponulywuoj—jayinu npuvyy ay} Uo apps of pasafjo svayzo pup sjpoo ouiddypyg fo s48a. Burwnaz3—TT WAV I, COx. 314 "Se ‘LTE ‘dd ‘alquy oy} Jo pues oy} ye MoT[oJ Soj}0UW 00M, RPS cee eae nee PoC hS | iip see Dea Thy | er pica Peay Poa | aT eC | it eta Repeal COS OE Pet ete eae OsGO TS eo Urea Maan |b So auae ian (ore ei eer | ope orn oo era LAY OL°L8 08 ‘OI 00°48 OF'% Mt RET} Sel 09°69 89°¢ c's ¢‘9T0'T 880 ‘On 8o°Ly T9TL 8 ‘SLT ‘LT 6. ere be oo [ese aad TP '89 88°% Gh '6 9L°6T €0°L OUOU AT[BOT}OVI | §°66 6120 0 16°0 6°ELT 6818 ‘T 8. Rae LPR esl eae oie 8 "SL 16 ‘1 cs'8 68 CT 1r'9 auou AT[BONOBIT | 8 ‘OL 06200 F0'T O'88L 8°880‘T LT 00 ‘00T 00°0 #0 FS 90°F €L°8 LI 8¢ C6 ‘FT 8°L G'18Z $980 ‘0 08 °T 8'CES 8 ‘0¢9'T 9L CS °L6 SFG 9F ‘98 02 °9 86 IL 90 °9F OF'6 0'8 $°48 82800 LUT $ GIs G'6PS cI na ids CASE | COMMAS SGA Sct TP PS 02 °% 6°L 9F “G9 8 ‘OL BL 6°29L 0620 “0 $0°T 8°L81 B82 ‘T PL 18°96 $9°S LZ 0G are 9L'9 G8 °TL 99 “ST 0'6 S‘10a ¢180'0 SLT GbOG P60 'T &L tea Seo TTTTTT] GB “GS 6L°% SL’¢ 8 "co La'8T Z's 0°69% 8080 ‘0 OTT @ ‘661 9 "P68 ‘T ae ek rae i Sane IL 8% 08 °§ 09°L 0g *¢9 $8 1% 9°% L968 $0800 80°T G‘961 8168 ‘T IL FL'86 96 'T CREA extras “Sigal | arcaes beak eas || poeiae $6 '6L 6°LT 6 °L96 9FE0 0 86° 02a ¥'089'T OT aah oa Saeco bao S| teas 06°% 9T°L 19 OS 8F°9 “OUON 8 ‘601 Z880 ‘0 98 'T 6 '9FG 6 'Sh9‘T 6 ay he cerca (ees Ta aly PS ‘8S 6L°T 68°L 86 TS 18 ‘9 8°IL SIL €980 0 08 ‘T T'S86 6 'Gh9'T 8 08 *F6 06 °¢ LOLS tr 0 ¥6'8 OFS 96 CT 9°88 L°891 €620°0 gO'T ‘68. €°L6 L GS "$9 80 ‘Téu 6F FS 70 62 °T 00 FT 8L "eT 8°68 T'8cs 60800 OL't | £006 0'80F ‘T 9 ae a Sanity | gtr aan ea SSOL GL 0z°0 PPT 89 °6L 98 ‘CT yuoU AT[BITYOVId| §°L0z 0920 °0 £60 891 G'GsL ‘T ¢ Rope eal | See aT LOA 940 “oT 66 0G QT “LT euoU AT[BOTYOBId) §*L6T L820 °0 060 9°99T € 666 id 00 ‘06 00 ‘OT LG°9L €1°0 61% Th 1% rd & 00a F'OST €hc0 0 L8°0 O°LST 0 °660‘T § FI G6 98°L 09 ‘FL VOU Neel ig) | Goeriaerr a recanaen OF SZ 09 “$1 €°66 @ 861 @820'0 00°T T‘Z8L 0°GLa‘T G 04°18 08 “CL 0G TL “ouON 00°T 0G °L6 &L°8L 1S 6 ‘96L LLZ0°0 660 P'6LT b S93 ‘T IT "quad IT "YUad bag "SO]L “SOUL “SODUM "SO]LM “SO]LM te 66 8G GF LV OS “moy 13d oyun | FONT | sroqauo SHEED) ERED, ieitract wo0qiBg ysV aINISTOW erdtienaticy pexty A GOaIaD 4 -yeay 1978 AA caromdag (Patan, [2 0} porsay ‘asnjal pus p ano On Oe USBUPIOAUNO | yse poy |—Jo coyouroop orwnbs Jog ‘JOHUI[O JO SISATBUY “yse Jo sIsAeuy : *poulnsuod » saTqysnqmo0p ['1Tg ‘d ves ‘[era4 JO uolWeinp pue ‘pasn 0}v13 Jo puly ‘[vl4} JO o}ep ‘[vOd Jo azIS [vIOIaWIWOD pu 9d1N0S 10,7 | *panulyU0j—yayuww DUD] ay7 Wo apps 1of pawaflo svayjzo pun sjpoo awddywyg fo sysaq burwna,g§— TI AIAVE 315 PHILIPPINE COALS AS FUEL. “STE ‘LTE ‘dd ‘afqvy oy} Jo pus ot} 4v MOT[OJ SoJOUOOT (ee 0°G08 GZ8T'T 169 ‘FIL x ¢ 016 ‘96 Pap mes eae me pine esp | as Mase ey | RS SRR AS teed ee IT aa | sg | ARS | Se ETT | OS 691. F'88s 69L1'L 90T ‘221 GROOO {SO a= i Roaer spain | cata ey aa | aa NET | gees celeb | Oa mea ac 002 OF 96 9°TIL 61 80 O°LL GSAT 'T TF0 ‘6 GTLO‘L QOL LENS re) ieemeremane SU3 Pe ss | pasteceeiegeg |lesar-an = aa SLL 61°16 9c FIL €°68 O8LT'T 18 ‘L G "9829 092 ‘9 82°28 OPT. 09°9 0476 1°61 62 86 01% LT GL 0'F6 LGLT'T 208 ‘OL “COL ‘8 C6L'8 69 'T8 &%'0 9° 08 ‘OL (v) 9§ 88 069% 9L LOT. Z'08 GPOL'T 610‘G ‘TIS 'F Lee ‘F 0¢ 18 cz0 46 °9 08 ‘TL (v) a8 86 8°cL cI LIT Z'88 CELL 'T £88 ‘6 "G86 ‘2 SIO ‘8 1818 90°0 8l'6 68°8 8ST 19°06 0TSL FL COL 8°16 OPLT 'T $90 ‘OT "89G ‘8 0098 G¢'T8 190 88 '9 96 ‘OT §°LT 84°88 G6LL &L cg 8°89 6PAL'T #66 ‘9 "ec6 ‘¢ 16 ‘G 8F G8 638 °0 084 18°¢ 60 #8 G8 0°62 ral 8 8°19 6FLI'T £98 ‘9 "888 'G gag ‘¢ Ch 8 eL°0 cok CLL (aad #618 9°98 ig 801 TLL SLLI'T 890 ‘6 T69 ‘2 0%L‘L €8 FS 000 16° 02 IT 0°62 08 06 0681 Me Tar 116 LQLT'T cog ‘6 "P80 ‘8 ASCH Nitin al pemcoeane mar al h ncuron onl 08's 10% 6°18 89 66 PLL 6 SIL 8°88 CPLI'T PSL ‘6 “062 ‘8 CGE Gs rursiecen anaes a #g°9 SOL Lp 10°96 QTL 8 GOL 9°16 6PLI'T 691 ‘L “C01 ‘9 GOL ‘9 LT'88 SFO 08°¢ OL IL 802 19 FG GLC L 86 CSL 69211 go ‘8 “PPB ‘9 0489 61°88 SF 0 66'S 60 TT 8°TF 096 0°31 9 801 G18 POLL 'T 618 G°LIs‘L QPS ‘2 68°28 620 88°¢ 6 ‘OT 0G 81°96 0'TS q OIL 1°88 PELL 'T 698 ‘L “004 ‘9 GZL‘9 06°28 oF 0 08'S OF IT 1°66 69 "C6 LIS i IL $98 OGLE TL eS 6 “ChO'8 G10°8 08 88 00°0 04°6 00°2 T 96% 90°26 6°98 (5 kas (an) PSLL'T CEP OT "G18 ‘8 806 ‘8 Te $8 IL'0 8g°9 00 ‘OL 0°68 14°96 1'6F z 821 o'96 6GLTT 88¢ ‘OT ¢°196 ‘8 066 8 T8 8 #90 GL°9 08 °6 0°68 0 ‘96 L°8¢ T ‘S02. 09 s¢ Lg 88 98 gs ts LL “asnyol *podojaasp *SIOWT *(N) (e0u0 *(O9) (0) *(700) ‘paumnsu0a 1810} 013 480} qamod “SOTIM. -1oayip &q) oprxouour wessxQ Oprltxorp “paArasqo AT[BNYOB UL [VLIO} BUT Jo -9s10y pozel |s‘padojoAop| “UuolyRI0 VO o00T udsOI}IN, uoqiBg woqiBp SB oHOUls poly se araysnd ‘ON Sepl[indg Jo raaod -dvao 4B pus WOIy Jo 3B [B09 9} JO -m100 9} oH 03B}]U0010g -as10H JO 10}0Bq | JUD[BAIND| 4juedieg asvjUsdI0g | JUATBAINbA {TatodvonDaT ane *(aumjoa Aq qued red ‘[Roo OY THOq 0} Poy 197UM DAIS SIOqUIMU) oSasvS ONY OY} Jo sasAjRUy ['Ilg -d ves ‘[el1} Jo uolWeINp pue ‘pasn o}e13 JO PUIY [VII] JO eyep ‘[vOd Jo ozIS [VLOIaMMOD puv sdINOS LOT] ‘paenulyw0op—jayunu Dpunyy ay? Uo ajps of pataf{o s1ayz0 pud sjpoo auiddrpryg fo 87897 Burwyaz9§— TT AAV I, COX. 316 “STs ‘LTE ‘dd ‘a[qe} om] Jo pus ay} ye MOTIOJ So}OU}00\T 9 00L 7B PUB WOT 10}vM JO UOTYBIOdBAD JUaTBAINDA Oo} P's | ASH “MS |} Z°0 SECO AE In leanne cua |ieaae eee me fale coe kee] ep || GIN te 9 eas O'GLL‘b | 0% 0D | Fg | ‘OIqVIRA | 6°T LFS G6 "8S GLO'L | GL6°L | PFO'L | 668°Ls | 9OT‘Ls | T6L°9s | G82"0 0960'S | 6T 0/2Ir | ‘a MNN | 0°0 FO ‘IS FIO"L | 09E°9 | 206° | Gc8"9 TIh'9 | GLL'G 810 ¢’c0z‘T | ST q | 2'OL “MSM | 0°0 68 Ze 0sz'L | 994°9 | 680°9 | 90T"L 1s9"9 | G86'¢ 1020 gces‘T | LT *(w110}s JapUNY}) oO | 6ST “MS 00 1°29 OF 2G e90'L | c6r’9 | OFO'S | TFo"9 TLp'S | 8ob'F 1220 L’TLE‘T | 9 q)9'6 |"MNA “AM! 0'0 86S OL ‘TS ges9 | 984°S | Shhh | O16"G 00F'S | 98h*F P61 ‘0 L°¥Go‘L | ST 0/32 “MS 0°0 GGL FGFS G90°8 | Sh8°L | OLB"L | SIS"L €IL'L | 289°9 £120 eeLe‘T | FT *(S[BAIOJUT FV UTVI PUB 9[ZzZEIp) I ‘O | 8 °ZT “ass T'83 | 3°26 GIS SF6'L | 869°L, | Gz" Th0"L c1g°9 | 00F'9 BCG '0 L‘LEh ‘TL | St “(cur d's[vasojur 4B spur Aysn3) 0 | 8 °9T “as 00 Teg 91°LE guge | cic‘ | Gho'G | STO"e |icez"h | OLP'> PCL'G 1666 ae OF ESS ee OLS) 0'0 9'TS 29 LE 600°9 | L6¢°¢ | 892°G | F26'h 98c"h | LISP 0st 0 G°L96 IL 29/96 |’ “MNN /|0°0 Z'0L ¥0'Th 696'°¢ | Ter°¢ | SPL“G | 068"¢ £06'P | 0c9"F 1810 L°L06‘T | OF *(S[BAIOJUT 4B Spurar Aysn3) O | g "pz “MS PT 9°6L OL OP ego'9 | G9g"¢ | 9FG'G | GLL'G 119'G | L08"¢ 0260 L’Gch‘T | 6 2‘O | 0'sT |'MAS “ASM! 0°0 £°69 GLSP 09E'9 | 986° | T99°G | FI6G LPL'S | G&P'G G1Z'0 9°068'T | 8 *(AqTBubs) 0 | g ‘Te “MSS | 1'9 Les 08 ‘0G F66°L | GST" | 8G0°L | 899°Z TLL "9 | 289°9 BE '0 Seer‘ | 2 q | 80 “M 00 GTS 6 68 coo9 | ca"¢ | LGT"G | Tra" 2720'S | 086 'F 8100 L0SI'T | 9 2 / TOL “M 00 0°89 98 6 geg*, | 828°9 | F89°9 | Geg’L 89¢°9 | 62F"9 L610 O'TLZ'T 1 ¢ OU GS AevSter 00 9°¢9 OF 1S 08%‘ | #ST"L | S00°L | L98°L 688'9 | P69°9 £020 €ole'l | F 9 | 0'@r “MSM | 0'0 Tul 98 °8¢ Z98°8 | PEO'S | FES*L | 109'8 86L°L | 199°L 6020 Pose ‘Ll | *(S[BAIOZUL 4B SpUTM AJsn5) O | g*sT “MSM | 0°0 TLL £6 "9¢ 09F'S | TIF | 900°L | e8T’s8 691'L | 0L6°9 0€z 0 €'06F T | % *(spurat Sysn3) 0 | 11% “MSM. || 6°S G'9L 66 "LE GELS | 199°L | 9FF'L | F68'8 9c6°L | OST’L EZ ‘0 ¥G0G‘T | T -mn0y : ez tL OL 69 v9 9 eee “uol} Sun} a ; SEE | citon) pore ‘9OBIIUS SUL a " A “St c SB ‘: L “au | 29H | Seay, | porminays| yeorwogo fos] “PO|rw09 10] 1a. | “wor | poww, | awou-tores | soy ‘9010 “Tru ur}-0uy Joes") “4m, uo Se) aw WED a eae 4804 FyIIOYZBIM IY} JO 9}BIS “qsoq -juad10d paseq jo ——___—_———— surmp eae uaa red ‘ON : Urey Irv JO Be ae AiO CTRIEER, —Jo O[TY tod “mmoy Idd so} Sulnp Aqrpramy | “yeztoq Jo. PUIM SUT[BAoId aBBIAAY | Cou [te -d cas ‘[elr pue ‘pasn 9}e13 JO puly ‘[ela} Jo oep ‘[eOd Jo azIS [BIO1aMIUIOD puv soINOS 10,7] “ponultyuonN—jayioue DjLUD IY 2Y} WO apps of pasaf/o ssayjo pun sjnod aiiddyryg fo 8189} Burwnaz9— TT Wav, 317 PHILIPPINE COALS AS FUEL. ‘Uol}e1odvas Joyveis. pue ‘ainjeiedwo} AOUMIIYD JOMO] ‘WOTISNqUIOD oja[duI0d a1oU SI 4[NSaa ey} pue ‘paiy-peeids puv poyoo-lules Ajo}eUIOI[e SYA 4] “YSIUY 0} J1e}S WOIJ PoqunjsIp Jou svM oyvaIS oy) UO yeoo 9y} ‘JaNJ sty} JO asn ay} ula “‘pery AT[eN}oV [vO ay} JO Jey} UPYY Sse, A[QeIApISUOD SI SISA[VUL [BOTMAYD 94} Aq UMOYS YSe ayy ‘e[]duies A1OVVIOGLT O}FBAINDV UB ULe}qO 0} Y[NOWIP ATOA SUM 4] “[vOd SIG) JO UNA oY} YIM poredutoo sv YSIY I1OJo1oy} 91eV API[IGe SUIMIIOJ-1oYUI[D pue yus}U0D Yse ofL “10A0 pois A[QeLepIsuod pu Peyvos Joy Used PEY II LOG [OUUN} oY} UL ULL] A]JUEpIAS peY 4 IOJ UA OY} JUaSaddea you saop 1809 SIM o ‘ *pasojour AT[BOUVYIOUL SVM [BOD BUIOS }eY] SUIMOYS 10}}BUI a[qIsNquIOD o[1}Z[OA Jo yUed aod T°g pue BIN}SIOUL JO }UaD Jad ¢°9 PouIe}UOD OSTe JAaYUI[O OL u ‘aiInssoid Wives MO] pUv diN}eJeduIe} ADUMIIYO YSIY sett}, ye pue UoT}eI1O0deAa ACT JEYMOTIOS V SBA Z[NSOL oY} PUB AV[NSe1 dIy sty deoey jou pynod oy ‘uo ynd seM UU MoU Y ‘“UoMMoIy UI UOljeIAVA [VUOSIEd OY} SMOYS 480} SITL wu ‘ooeuIny of} JO Loop oy} WoLy Surddoup ur sooord 0} s[[eJ YOM JoyUL[O JUaTdIoUT yJos B Jo ued aed F noqy ,; ‘ayouls jo uolonpoid a[qviepisuoo B INOYPIM Sul}jes INO YA [VO URI[VIISNY SIY} UANG 0} a[qIssoduit st 41 yey} SOJBIIPUL FOIYM ‘WSIY SULBUIOL [[I}s o8vJUeDI0d oY} ‘ayouIs SS8] Soonpoid sAVATe YOM [BOO AuIN] St SIq} }eq} JoRZ OY} Jo ojIds uz “OMOUS JO UOTJVULIOZ 9} JUOADIG 0} OPVUL SLA JLOJJO [VUOT]daDXO 4So} SI} UL ~“ASu0ue Feo JO SSOT OWIOS SB [OM SB 9YOUIS Sso[pooU A[Pe}qhopun St a1aq} S}soe} asoy} JO YSOUL UL PU YSIY [vod J0}}vOS 0} ST USWEIY dalzVU JO AOUApUE} OYE x ‘spunod QZ‘ JO U0} sad Gy 'OTa Je JUSUIUIOAOD [IAIN OY} JO Sprvd pue spoys [vod oy} UL polid “e[IUeW Ul poseAT[ap adIad JovAWUOD v ye SPS ‘ON Tesodoig avjnoi19 uo paseyoind sea 4J ,,,001M} payord puke poueosos-afqnop,, Si jr ‘“A[ddng jo neoang oy} Aq peystuiny [ONJ A1OJVIOGV] 9} SI SIMAL ¢ ‘UIVI=1 / AYS JSBoloAO—o + AYS ApNo[o—o + ANS on[IG=—q _; ‘TeOd OY} JO ONTVA GILO[vO oY} PUL O1V¥I [ON OY} PUB UOqIvd sind JO aN[VA OYTIO[Vd oY} ‘TOYUI[D oY} puL YS oy} JO UOLzISodtmoo oY} Woy payepno[e9 1 ‘damodasiog 7 spenbe "9g .0OL 7% pue WoOIJ InoYy tod payerodead 19}eM Jo SO[TIY G9°GT 3 ‘o7 J@ WBO}S OJUL pasuvyo SI of oIN}eIEedme, AUB Ye PINDIT ey} FOIA YSno1y} ‘wva}s poyeanjzes JO JRoy Jue] OYF=~4 LOY PUY of OANJBIOdWI9} AUB YY Uva}s OJUL pesuvyo St ,0 }¥ PINDIT oy} YoY YSno1y} wve}s poyeinjyes Jo jeay [e}0} oyJ=Y osoyM CUULUTPEYTIM) 921F1800000'0+<1LFLZ800'0—22L64'0—S 68gG=4 (8G ‘80G 6 ‘(O88T) “Wuy “Palm “y ‘UUeMITEYUTA) 726 F800000'0+.1LF6TE00'0—1820L'0+6'68S=V CGE9 “18 (LEST) “perVd op ‘wary ‘yIneuzey A) 3¢08'°0+G°909=V :peseq SI safqe} []@ Jo JAvd Jo}voIS 94} YOLYM UO X[NWMAIOJ SULMOT[OJ oy} wWtOIy poayepnopeo oq AvUI AVL ‘siesursue Aq pojdedov A[[VioUes ov TOIT «W948 poyBAN}eS Jo sojsodorg oy} JO sarqeL,, SApoqved Woy woyr} o1v sAOqUINU dS9y} SOMOTUAAMOD JOY ‘WE}S JO YvoY 4U0}L] OY} SY UEYL} Suleq soruo[vo G'9g¢ 82—6'8¢9 ‘19}VM JO SOTA OY} TIIM JOqJeSO} osoL “SZ Sf 19}VM pooy oy} JO yVY} PuUe G'SG9 SI (‘UI “bs Jed ‘sql G GOT) sejoMIIZUOD OLeNDs 10d smeASO[Iy OTF) 1 wILa}s poazeinjzes of} JO “H .0 #B JoJBM WoIJ Sot1o[vo UL yvoy [¥}O} oY “Ppodeplsuod eq }SNUI 10yeM poo; OY} Jo ounyeredme} oY} pu ounssoid wivo}s OU, y ‘SpOYJOUL PAVPULIS 0} SUIpP1OdIe OLOMAATA “WW “II AQ pozATvUy o ‘T[VA aSpIIq oY} JOAO pot1Vd YSe oY} OpN[OUI JOU sa0p SIqL p ‘SISATVUB 9}VULIXOId 9Y} WOT poze[NI[VD o ‘TH ‘8 ‘(LOGT) “V 998 1DULNoP s1yT ‘XO JO poyOUL oy] 109Je IOMTEM “S “H “AW Aq-pezsTvue ANSON q *(Arno19Ul UT seyoul OF) (Ul ‘bs Jod ‘sql 4°FL) ‘Wo ‘bs ted swumvISO[IY FEO'T OF PeNbo A[WA0JIUN Sv UOYye) SBA dINssord DIJOMOIeG OL v G'98G ‘SOIT SE¢‘OL=G'196'8 X (WOT}e0dvAe Jo 10}0R7) sv oinsserd of1oydsomye 7@ "D ,OOT 7® pue Woy JUETRAINbs 94} BATS ‘o[10q ay} 0} pay ‘e'TOGS “9IS-I1€ ‘dd ‘Il AIAVL OF SULONLOOK 2 75034 COX. 318 “LOYJVIM OY} SU[PIVSel VIP po[lejep ey} OJ NvaiNg 1ayIVeAA EY} 0} Pe}qapur me J “BUI}vOyAIOdNS JO JUNOWL SIY} Aq PoAvsS svM JON} [VUOT}IPpe yued 10d TI ‘@ Ul UOATS Sony[BVA ay} Susy “SUIYVaYyLOdNS ZZ SVM O19q} [ROD UVI[VIJSNY JO }So} IloY} UT ‘AJOTIVA OWLS oY} Jo o[dures eSsva0eAv AW uevy} JoysIy Jue. Jed OT pur o[duivs pajoojes AW uty Toysiy Jusd Jed fF Sl YOIYM ‘VITBIJsNy ‘PUOSP[VAJSeM WoAy [vod IOJ OO'S MONS ‘AuvdmoN yYSIT pue prosrey olszOeLM v[IUe Oy} JO Byep 8 z ‘Jeyeoysedns vB JO asn oY} Aq ON[VA SIY} 0} Ppaonpad SVM oinj}e1sdu9} OL £ ‘pomva daIsop OY} 0} Surjyvoysedns oy} Aq poavs SVM [aNjJ [VUOT}IPpe Jo JUed Jed FG ‘san, vA ssoq} Susy “‘yeoqdedns (“\f .00Z) “0 oT LT 4103 GLO puv yeoqaedns ("\ .O0T) ‘0 GoGG 1oy Goo AjeyeUIxXordde sy yf yey} “QueJsuOd jou st UIve}s poyvoysedns Jo yvoey oyToads oY} }Vy} UMOYS sey uoNeSNseaul yuadoy ‘you}s aq} Wsno14} JSOT SBA JVOT SIT} S}S9} SACOG" OY} UI Joy ‘Suyyvoysedns ou Jo UoTduINSse oY} UO peseq ov S}[NSOL pu SUOTJV[NO[VO SUTMOT[OJ oy, ‘Suryeoysedns ¢*,cy x “4SBOO VUIYO OY} WO1J JOYIVUT VILULIN OY} YOVoL ALU YOIYM [BOD oY} Jo AzITEND oy} AVM [e18UeS B UT JUeSeIded [[IM PUL BdINOS SIq} MOAT 1809 JO Vopl Ley V AIS [[TA VJVpP asaq} JLY} JYSNOY} St }f “IOSVOUT S] 41 OF PAvSo1 YIM UOLVUILOJUL INO pu [VOD aSeUTYD JO 4S0} B oYvUT 0} O[qe Waeq Jou oALY I 8901S ING “BJVp JOOIIP SB AIOJVISIVVS Sv JOAGU O1V BdUOLOTIP AQ pourejqo ve “G LoquIedeq JO 480} OY} JO SISvq oY} UO pasn [vod ULITeLIsNY Jo JUNOWL aq} Joy Suyjoed100 “F Jequiadaq WO 489} ey} JO yaVd [eUOT}LOdo1d oY} OYE} 0} [VOd BSOTIYD OY} 0} ALVF ysvoT 7" ST qT ‘[Vod 10}0q & YIM poXIUL UEYA J0}Jeq suanq Ud}JO SUOT}IPUOD [BoISAyd 0} SUMO [VOD 1OOd v 9dUIS pUw “[eNJ JO SpUIY OM} 9Y} JO JoMOd OAT}VIOdBAS SAT}LIOI ay} OUTUILOJAP 0} SEA S}S0} OSAaq} JO Joelqo oy} edUIS "pasn SBA [BOO ULITVYSNY YSIVI}s ‘peo, SurusAd oy} Jo Yvod oy} Sulinp yey} puv sory oy} Suruvelo uayM Sao]10q peyuLg oY} JO aU WoOIy ddULISISSE jo3 0} Arvssa00U SVM JI OINJXIUI 94} SUISN UoyM yey} Juoder Ady, ‘ouo[e [vod uUvITVIISNY UO apeu SVM 4s9} OATJeIOdvA® ANOY-FZ ¥B aIqISsod sv suo) -IpUo0d [VOTJUep! ATIVSU SB JopuN AVP ZUTMOT[OJ oY “ATEAT}OEdSeL ‘SOTIY ZOF‘OE PUP OGL‘GT SUISN [vod UITRIISNY PUL oSaUIYO JO eANj}XIW v YIM UOT}IpuOD Buljeiedo ivjnSei JepuNn sJe]10q Jey} UO }so} ATWVAOdvAD INOY-FZ V opvUL AUVdMOH WYRIT pur prodpIeY dsjOoT VIIULI eu} ‘906T ‘FP Jaqme0eq UO « junowe sq} Aq JeSivl e10Jor9y} st poyesiodvae ATJUeIvdde JayeM oY} PUL ‘10J Pa}UNODDvUN }VYy] PUB UOIVIpPel 0} ONp sso] oy} Sse_ oY} JULId 94} saZIv] ey} }ey} UMOoYS sey sduetedx@ "juod sod May v AQ pasvedour oq A[qeqoid p[noys Aoey} SIq} YIM peredurod ere suoquinu Suipsooid 9y} Ua a ‘OOVJANS SUI}VOY IaJVM JO SdojJaUIIDep aIeNbs OSG‘st n ‘aovds 11v jueo dod FOF ‘oovVjans o}v13 JO siojoMOep aieNDs 6'ZGF 4 “ZS ‘8 (906T) Maraaa ULIISDY WWJ “YAL[_ 9d] Pue ISVIOIS POH AV[NSUT oY} }e ope svM UOSTIvduIod Jo sesodind 10J poajze[No[Vo v1eM VJEP ISeq} YOIYM WoO S80} OL s ‘UOL! OY} JO UOl}EPIXO 9} 0} ONP ONIVA JAI}VSIU B SBATS A[[VNIOR SIV s “(qu90 Jed QT Japun) eyouIs Ou AT[ROI}0RIg b 319 L. Dy PHILIPPINE COALS AS FUE "Gas ‘Lag dd wo 9[qv} oY} MOT[OJ sejouOOT nieces an 8642 ee 13 cecmeimelULGRS TTT OFS SL |] 99452 J | 99K L = | OTAYSNqQuLOd Jo yrun T Jo onTVAa yeay [BIO], 600 Teva irises sien "ese erantats| rye an ee ee ace ood |W emcee OO Te sie | ference! 500 Tet | aes 100 Tes | caries [ist pp eam se pa eee as ree 1810.1 QGW LEE y SCG Gira | asses pees | aieny an aad Ne ae ace | reo Gh 6 88h LL6E | 98G‘T =| TG"eL =| Goa‘T = | payetnoyea oq Avur Loy} YoOryA\ UrOIy poureyqo 1B BLRp Jr paztutejt Ajaywredos oq ABUT SaSso[ aso} JO aUIOS ‘10F poJUNODDBUN PUB ONBIPBI ‘ ( i 0} ‘ITV OY} UL sINysTouL 9y} Suryvay oj ‘suoqied -o1pAy pus udssoipAy powinsuoosun 0} aNp ssoy *L OFF Shs 68'S FIs 15202 GOS 6G 1&@ 66S 966 86° COS Se |eooeenaennnns OSNfal pUB YSB UL 9[qYSNqULOd OF Ol sso'T “9 18°T OFT cet e0L F9'T cst 00°C 0 gge> oP 93'S Qa FR eUOGAV JO UOTJSNG U0) VjJaTdULODUT 0} ONp sso'T *g St'e— | 420‘T ¢9 “CT 986 ‘T 0L ‘ST 066 ‘T GC "9% NO RS Tae Gate | (ae Pos ero Coo ea psosus Aoumiyo Ap Ur ABAB palliBd yway 0} ONp ssoT “F 18% VOC eee ral Necie Rew | are aera Pee Sete Geren ae ¥0°% 291 02°% TLL 02% LOA Re keen eer eee ena [page pas SCS oUasOIpA jo Suruing oy} Aq pauroy oINASIOUL 0} ONp ssoT “g 610 cL 80 6. 820 61 610 CT 180 iva 180 BGs | Caco aga ESTE [BOO OY} UL OLNYSIOU 07 AUP SSOT °Z ec'6e | 080's & OP 860 ‘P OF Ig 10a 'F 98 °8¢ GOin |) as. ams | GSE ieee via[tog ayy Aq poqiosqu yw “T 2308 *SOLI “qua0 ‘SOIL “quad “SoTL “quod ‘SOLE qua0 “SoTL “quad “Sold ‘10}0 BT ad -O[8D lod -O[BD dod 0) 1:70) lod -O[BO Jog -O[8D lod -O[BD I -dunT *yoR[s pur dun] | ‘yoxls puv duny ‘dunt peyoapeas | yovys puw dummy | yowps pum CUT po OZIS [BIOIOUIULOD “puvys] nysuLy : (s}BiBy) TUBy -ourysox ‘under “AOTIBA. MOZYOVT ‘VIpBIsSNYy “AOTIBVA MOZYOLT ‘BI[BIISNY *puaspe -4SOM ‘BITBIISNY *puas[Bar -1S9.\4 ‘BI[RIISUY ' “puos[BAL -4S0\\ “BITRIISNV [809 ay} JO doIN0g 9 o Pee ap ae Po RES 489} JO ‘ON ‘a7quusnqwoo ay, fo anjna Burynay ayy fo woung.ysip 10 20uUn,Dq JD9H— TIT ATA, COX. 320 ‘Gse ‘Tze “dd wo a[qv} oY} MO[[OJ saj}0Uj}00\T OPnt tad Poi Lia|ae as oa S00 [ac)| fincas nea ‘OOT 60 "01 Og GG'ST | $96 60'6I | o48‘T | e2'8z | Gz0‘% | O8'se | 9TP‘s | GO'TE | Sha‘s | 64'°% | I88‘T |-—~ pore(Morvo oq AvuT AOI YOIYA Woy poure}qo d1B BBP JL pazruajt Ajoyeredas oq ABUT SOSSO| asay} JO OUIOS {10} poJUNODOBUN PUB UORIpPLI 0} ‘118 OY) UL oINyslOUL VY} SuIyway 07 ‘suOqiBd -o1pAy pus uasorpAY pautnsuodun 0} anp ssoT *2 GF IT 828 99°FE | SFO‘T | 90°8T | 660'T | 04°6 #89 G8 ‘FP S1€ 66 °E O8z 66°C Os fies rere dSHJoI PUB YS¥B UT O[GSNquUIOd 0} oNp sso'yT *y 90°S GGG 88°L 96S #0'¢ 69S 00°0 COM ane si pares Ue Hersam. oo | he Decree ees orang 96 'T 8ST —-=> ,u0qiBd JO MOTSNGUIOD ojJo{dULOdUT 0} aNp ssoT ‘G ¢6 61 PPP'T | GER] | O04‘ | GucZE | 9L0‘E | LO"LE | 2Os‘T | O8"9T | OT@‘T | GP‘ST | 9FE'E | E9°9E | ORL ft ag eg a psasea Aouuryo AIp ul AVMB poatIVd Ywoy OF ONp ssoT ‘p £9°G T6T L8°T S&T 6L°T as 6L°% L6L 9¢°% 98ST CoG €8L $8 '% GB [6h =| cinas eyoctaee ah gee aa ee pee » uasdOIpAY jo Suruinqg ayy Aq poo} a1njstouL 0} oNnp ssoT ‘¢ 04°0 1S 99°0 LY c9'0 LY ¢9°0 OF 08 0 GG 6z'0 IZ cL'0 Clits ae tera gaa q [80d OY} UL INjSIOUL 0} aNp SsOT *% GE'z¢ | zuL‘s | ons | 060'% | c9°LE | 10L‘G | FO'Ih | 68's | GL°ch | 8608 | GL‘Eh | GE'S | 08'0S | 090'F | vJo[1og ayy Aq paqiosqe yw ‘T "yu90 “SOL “yu00 “Sold quo0 “Soll “qud0 “Solr | “quo00, “SOL “qud90 “SOLL “quoo “SOLL “s10108, lad -O[R) lod -O[BD Iod -O[B,) lad -O[8D ad -O[B8D lad -O[BD iad -O[8D 19 BT ‘dunt Yovys pus duny]|yxovyjs pus duny|Hovys pus duuny, ‘dunt 0} vag ‘dunt 0} vod S CUNTICTE ete: | eamcgatanpena aaah? esp RakeL ep ah Se ee OZIS [BIOOUIUIO() “Pp ON ulvas‘UOl] | ‘WOTBAIOSAyY “MOTIBAIOSAY “UOlBAIOSAY 0 ; f “QOUTAOIg -RAIosoy AIBN) — ATBITTI AIBN ATRYN eee eee GPO BOIS | [vod ayy JO don0g ‘pus uve ‘purls] UBeg | ‘puURIST UBVeg | ‘puUR[SsyT UBIYE a a lieqny ‘uedey SI er Ir Or 6 8 tee gOl gtr renamn ARATE RIL Ge ieee 480} JO ‘ON *ponuryuoj—a/quisnqwwoa ay. fo anyou buryoay ayy fo worngitjsip 0 aoun]nQ JOAH—TIT ATV, 321 FUEL. AS NE COALS PHILIPPI SLOJOVJ AUVUL OLB 919T} VY} JOVI VY} OJ SB [[OM SB “WOSBaI SI] IOJ “LOL19 a[QVAOPISUOD 0} a[qRI[ o1e SaSeS AUMIIYD oY} JO sarnjyesadute} oq} JO Sulpeal aq} pue sul[dues oq} se ‘ayeurxoidde Ayuo st on[ea sig *(j—L) FS OXPIGNSNGmod Jo iYSIaM JUN Jed sasvs ony 94} JO JYSIOM 94} —SalIO[vd UI SSO] SITL p 1809 Arp of} UL UeSOIpAY JO YSIOM Aq Jed [eUoTsodosd oT} st H oLoMM [(00T—L) 8F'0+9'989+2—00T] HE=Sersoy¥o UI SSO] SIL o ‘sose3 ong of} JO oinjzesedule} 0y} L puv ainjvaeduie} WMoo1-a1g 9} 2 S aTqlsnquLop aq} 07 Ppeilejet einjsTour Jo yuad red 94} SI M 91eyYM [(0O0T—.L) 8t'0+9'9e¢+7—00T] M=Sels0[¥9 UI SSOl ay “ATWO IN}S[OU dIdoosorsAY 9} 0} SIOJOI SIMD q "G'9E8G X0.00T 1% pue Woy pa}e1odvaAd 10jeM 9} —=—SOlLO[VO UL ANTVA SITL v Besoin | eencac ad [imac BEEZ FT| 908% || 80662 | ==] 06819 | AH-] GOR9 [om | gea‘2 |-7=-=F> OfaTasnquod Jo 4rUN T Jo NIA way [e10g, cre [ara | es eee pea (ee SS sees |e 0) Te |e LOE |p 00/1 |e HUGH ieee eRe aan CRUDE DT rao ae en pa ea 18}0L, PEL] Sd braille hte nea wala sea PO GaGa ecle Ae Ae. Ht TGs 86°G | GBBT | €8°6 TT4 ~~ paze[noayBo oq Avur AoY} YOIYM WOIF PoUTYy -(0 O18 BBP JI pozturayt ATa}BIBdes aq ABUT SassOoT ; asoq} JO 9UIOS ‘10J payuNodDDBUN PUB UONRIPRI 0} ‘ITB OY} UL BINQsTOU ay] SuNBay 0} ‘suOqIBd -o1pAY puv usso1psAy pauinsuodun 0} @Np ssoT “1 SAS a |S oAe 09°s c9% 16% $91 IL T S21 FOIL | FFL 89°9 PLP 686 G19 "> esnjor PUB YSB UT a[qisNquULod 0] oNp ssoT “9 reise (te Moaegsl hadeaear eG [ch ee ae Moana aol ea | GED, gig 90'T 89 90°T 99 1g°0 Xe ~~~ stloqrBd JO UONsNquLOD 939[duOdUT 0} ONp ssoT “G rene a | oak T lone a al lia Te | 80°8I | S0S‘T | €8°GE | ZIO‘L | LPL | P16 LONG AMG G licen a a aS eee BS ee Les -foumryo AIP UL ABAB padlIIBd yBdY 0} ONP ssoT “fF (Pree ica sees es €6°T | PL PL'% | 861 18°% | LOZ LPL | $6 ere |} 20, CHC ieee (7/0 an | mem enceaneLL.JOLD ALT jo suruing oq} Aq poof BINJSIOUL OJ ONp SssoT ‘g ST Reacaaeea lls Cee 9F 0 ia SLT 68 @a'T 88 e's 00% G6'% -| IST 890 GF fg [BOO OYF UL DINISTOUL OF ANP SsoT °Z ay GaSe as PRG chee) G68 | 0468 | FO'TS | 8490's | 68'zo | aIs‘e | OF-co | sFS‘e | OL'IG | TAT‘s | Fa'EG | ea6‘g |--—~77----~--—----—---- paaqtog ou) Aq poqrosqe ywoH “T "yuo0 “SoTL *‘quo0 “SOL “qu00 ‘SOTL "yue0 “SOLL “qu00 “Soll “qyuo0 “SOL “yu90 “SoTL *S10108. dd -O[BD log O[Rp Idd -O[BO) lod -0[BD log 50) 78) 1d -O[8D iad -O[RD 708 a aan 2 === === === === ‘dunt -dunT ‘dunt YOVBIS pus dun 7) xovjs puv duny ‘dummy Deporte: Starae nee sho sr eaO2 18 'TBTOLOULUION), Re PON “AB : A ( : “siog “slog LH BIS NA(0) GRY Nes | as ee “uo ‘Bury OTMOd IsuBUIOD ‘nqag| ‘IsuBtOD ‘nqaD ‘Duds uyeg | ‘puvysy ueyeg |-rosay Are TTI [B00 aq JO oIN0g G ‘puxis, ueyeg O@ 6I ST 20 9T SI Lidl LNs Ml IPisehane Mahia pao win aT Ane wees aRS yp aural e 489} JO “ON Cox. N N ae) ‘a[qisnquiod oy} UI Woqavo Jo yYSIEM Aq yAed [eUOT}IOdo1d oy} OD pue ‘sass eny oY} UI eUIN[OA Aq ued aad ay} ere *OOD pue OD e1ojoq se pue (88z pure egg ‘ze ‘g ‘(ZSST) WesuNGonsi0j}UQ eYOS[Weqo-oMIEyL ‘WesMOYL “f JO SLoqINU oY} WoIJ peye[No[Bo) epIXOIp WoqiBo 0} eprxouoUr uoqivd Ut peule}u0o uoqivd JO WYSIOM jun euo BSuluinq Aq pejyersues sotiojeo jo Jequinu ey} St GOL‘G Ayyuenb oy} a10qM 9 x20 OP 8 5 hours —---=|" 7.312 104 380 5} hours ==--| 8. 226 117 385 53 hours ___-| 7.875 | 112 395 5} hours -__.| 8.507 121 430 Gibours= ses mar7eS 75m panel 2 a |e J Total ___/197. 640 Average| 7.906 | | PHILIPPINE COALS AS FUEL. Bt TaBLE 1V.—Observations in detail of the tests of coals—Continued. E.—SECOND TEST OF COAL FROM LICHZOW VALLEY, AUSTRALIA—60 FIRINGS DURING 6%-HOUR TEST. [Test No. 5, Table II.] : Steam pressure Oe oie Hees of coal Eons Fire di gauge. oa gases, inpercent,| PUrmed— boiler— raked Gleact Time after pure of sue aa starting: se Pounds] gases, | Dur- Dur- time ees square sates base oll Co. | O2 | CO ne Total. we Total., Stat ing. THANE. inch. riod. riod. ing. Beginning _| 7.031) 100 320 shoursss 2 7.101 | 101 329 hours 7.453 | 106 389 hours 7.593 108 439 1 hour ______ 7.945 | 118 427 | | 12 hours ____| 8.156 116 425 1} hours____| 7.312 | 104 | 485) '11.8 | 3.4 13 hours ____| 7.242 | 103 445 2 hours -___- 7.242 | 103 470 || 21 hours ----| 7.804 itil 517 21 hours ____| 8.015 114 507 22 hours ____| 8.929 | 127 420 3 hours _____ 7.734 | 110 395 32 hours____| 8.015] 114 400 33 hours -___| 7.523 | 107 397 3? hours____| 7.453 | 106 395 4 hours'___—_ TTL | ab) 400 | 42 hours ____|_ 7.523 107 407 | 43} hours ___-| 7, 453 106 400 43 hours ---_| 7.242] 108 og \10.2| 8.0 5 hours _____ 7,664 | 109 365 52 hours ____| 7.528 107 360 | 54 hours -_-_| 7.734] 110 348 | 5} hours --__| 7.453 | 106 364 | 6 hours _____ 7.945 | 118 340 6 hours __--| 7.101 | 101 339 | 63 hours __-_| 6.961 Qf) eeenes| J SO elt 35 ge |e 00 am et 200 ba | eee | eee 6} hours -_-- Total ___'211. 982 : f Average) 7.571 328 COX. Taste 1V.—Observations in detail of the tests of coals—Continued. F.—TEST OF COAL FROM YOSHINOTANI (KARATSU), KIUSHU ISLAND, JAPAN— 61 FIRINGS DURING 7-HOUR TEST. [Test No. 6, Table II.] Average com- « Kilos of Steam pressure Postion of flue Biles oCont water fed to | Fire gauge. Tem- |gases, inpercent.| ?UtReC— boiler— raked) (jeg n- ee ue ed fire ‘ ure 0 slie- |"p0 BE | ene ine ea, | Se StaTuIn Ee. per |Pounds| gases, Dur- Dur- time | Start per |base of) ing ing after| ~~ square square| stack. CO. | Os | CO pe- Total. pe. Total. start-| 128- meter, | 12ch. riod. riod. ing. Beginning__| 7.804 111 330 |) 2 hour __--- 8.015} 114 | 340 | | hours 8.015} 114 333 } hour ___--- 7.945 | 113 | 320 | | whoure==—= 7. 734 110 270 14 hours___| 7.875 | 112 | 335 | 13 hours____| 7.523] 107 359 | $11.2 | 5.4 | 0.6 |{ 60 380 | 800 | 1,450 |___-:_|----__- 1g hours____| 7.875 | 12 | 325 | 60 440 | 300 2 hours ____- 7.875 | 112 338 | 60 500 300 21 hours --._| 7.875 | 112 385 60 560 | 300 2} hours --__| 8.015 114 360 60 620 | 300 2% hours _--_|" 7.523 | 107 335 60 680 | 300 3 hours _____ 7. 453 106 310 60 740 0 32 hours --_-| 8.015 114 325 ( 60 800 |} 250 | 3,200 33 hours -_.-| 7.593} 108 | 33 hours ____| 8.015 114 | 4 hours _____ 7.945 113 j 5 42 hours -_-_| 7.593 108 | 43 hours __-_| 7.945. 113 4} hours -_--| 7.875 112 5 hours ____- 7.875 112 5i hours ___-| 8.367 119 53 hours ____| 7.734 | 110 [ 5} hours -.--| 7.875 112 | 6 hours _____ 7.593 | 108 5 63 hours ___..| 7.523 107 1 63 hours -_--| 7.875 112 6} hours __-_| 8. 086 115 7 hours ____- 7. 664 109 Total ___|227.100 | 3,230 Average} 7.8381 111.4 PHILIPPINE COALS AS FUEL. 329 TasLE 1V.—Observations in detail of the tests of coals—Continued. G.—FIRST TEST OF COAL FROM YUBARI (HOKKAIDO PROVINCE) JAPAN—59 FIRINGS DURING 5-HOUR TEST. [Test No. 7, Table II.] Steam pressure SOOT ROnHTNe Kilos of coal Eilosol Fire gauge. Tem-|gases, in percent, PUrmed— boiler— Taked) qjoqy_ pera Or’ ledfire, mimeraiter ture of: slic- time § a flue ed, Starting. ee Pounds] gases, Dur- Dur- time atten sabes eaare Baseiot €O2 | Oz | CO res Total. pe Total.) ars ing. AOR, inch. riod. riod. ing. Beginning _| 8.226 117 335 111 354 113 360 115 354 | | 120 330 1+ houts ____| 8.086 115 380 )E Uhours.._| 7.588] 108 | 854) 1 9 4/55) 9.4/1 5 | 340 | 925) 1,950 | | 1}? hours __--| 7.875 112 399 55 395 2 hours —---_ Hela) 102 360 30 425 21 hours _--_| 7.312 104 370 | | 60 485 2% hours —_-_]| 7.312 104 364 60 545 23 hours____| 7.031) 100 | 410 | 55 | 600 3 hours _____ 7.242 | 108 460 | J 55 655 82 hours ____} 7.812 104 438 |) 55 7 31 hours ____| 7.804 11 445, | 55 765 3} hours _-__! 8.156 116 448 | 55 820 4 hours _____ 7.875 112 454 55 875 42 hours --_-| 7.804 111 460 fp SO EO O87 5d 930 41 hours ____| 7.382 105 447 | 53 985 43 hours ----| 8.015 114 420 | 55 | 1,040 5 hours --___ 8.297) 118 377 |) W325] (1X07 259|/0 = 526 lh Gy102)|eeemes 5 00 Moye NGPA, TAS || PA BSNS) |} GL) | It O7.259| See Heal 02) ears | ee paaaee Average| 7.750 1,7) SER oe |-----|----- 68) Glee 305) | =aeaess | Saees ee iu 330 Ox. TaBLE 1V.—Observations in detail of the tests of coals—Continued. H.—FIRST TEST OF COAL FROM LABUAN, BORNEO—112 FIRINGS DURING 7-HOUR TEST. [Test No. 8, Table II.] Steam pressure | MOTE Aue Kilos of coal | \. ene to | Fire | gauge. Tem- |gases, in percent,| PUred— boiler— raked) Gjogn- } eee | che. \edsire,| Time after E lie eq, | me EAS. a Pounds gases, Dur- Dur- time | After square) snare letack,| 002 | O2 | CO| pee | Total.| TE | Total. iter ing. meter, | 12ch. riod, riod. ing. Beginning__| 7.664 109 875 |) ( 2 hour _____- 7.945 | 113 | 340 | | 4 hour -_-__- 7.172 102 330 3 hour -____- 7.523'| 107 | 340 | | ivhours=—ae 7.382 | 105 380] $11.8 | 6.0 |----- 4 1 hours ---| 7.172 | 102 | 375 | | 14 hours -_--| 7.312 104 350 1g hours ...| 7.312] 104 | 345 | | 2 hours ____- 5. 976 85 280 | J l 21 hours ____| 6.609 94 300 2} hours ---_| 6.258 89 | 365 2? hours ____| 5.484 78 315 | 3 hours -___-] 5.273 75 235 31 hours ____| 5. 625 80 320 | 31 hours ____| 5.344 7 $25 | r11.5 | 3.5 |_--__ a) 6) 3} hours ___-| 5.625 80 400 | 75 4 hours _____ 5.765 82 400 | 75 | 42 hours _-__| 6.328 90 410 75 | 43 hours ____| 6.609 94 415 73 43 hours____| 7.242 103 400 | 75 5ihours 7.593 | 108 425 75 52 hours __-_} 7.312 104 410 || 75 5: hours ____| 7.172] 102 400 | | | 70 5} hours _-__| 7.172 | 102 410 | esis : 70 6 hours _____ 7.312 | 104 | 9450 ; S=aliparay 70 | 6} hours ____| 7.804 111 a 495 | pe ao 64 hours _---| 7.312 104 9510 / 70 6} hours ____| 7.593 108 365 61 7 hours _____ 7.312 | 104 370 |J lL -@ Total ___|198.202 | 2,819 10,935 |_______|_____ pio) Average| 6.8384 CYGPAN BIT | se 64 | ® This coal was unusually sooty, depositing enough on the tubes in this day’s run to burn off at this time. > This group represents the extravagance of a native fireman. Green coal was thrown onto the fire and then mixed with that already on the grate, causing much loss of fuel. PHILIPPINE COALS AS FUEL. 331 TaBLeE [V.—Observations in detail of the tests of coals—Continued. I.— SECOND TEST OF COAL FROM LABUAN, BORNEO—113 FIRINGS DURING 63-HOUR TEST. [Test No. 9, Table II.] Steam pressure pouuonobaue Kilos of coal | los oe 9 | Fire gauge. Tem- |gases,in percent. PU™ed— boiler— raked) qyogp_ Peres vr led fire Time after : puree | sic: time starting. uc Pounds) gases, Dur- Dur- time ater: sane square paseo CO. | Oz | CO pe Total. pe Total. seed ing. meter, | 12ch. riod. riod. | ing. Beginning__| 7.804 | 111 340 (ees + hour __-__- 7.664 | © 109 340 | ia) + hour _—---- 1. VWi2 102 335 75 3} hour —_____ 7.172 102 350 | 65 ah In@ybyP —— 7.523 107 350 65 1: hours ____| 7.593 108 375 65 12 hours _-__| 7.172 102 880 | +11.8 a6 aeone 4 65 1} hours --__| 7.172 102 350 | 65 2 hours ____- 6.961 99 395 65 2hours_.-| 7.172| 102 | 390 | 70 23 hours -=--| 7.172] 102 | 2485 | | 70 2} hours ----| 7.172 | 102 | »485 70 3 hours _____ 6.961 99 | 9500 | 70 31 hours ___-| 6.820 97 405 |) 70 34 hours ----| 6.820 97 895, 70 3? hours ----| 7.523 | 107 410 | 70 4 hours _-___ 7.523 | 107 410 | 70 4} hours —__- 7.875 112 440 70 4; hours ----| 7.523 | 107 3875 | 70 43 hours ----| 7.875 | 112 BIO ep auligey lees 70 5 hours -__-- > 7.382 105 400 Are apuan 70 5} hours ____| 7, 664 109 . 375 H 70 53 hours ----| 7.382 105 385 | 1 70 5} hours _---| 7. 453 106 390 70 6 hours —___- 7.242 103 405 70 6} hours __--| 7.664 109 465 70 61 hours ----| 7.593 | 108 460 |J 61 63 hours -=_-| 7.172 BG) ese [need una Ly 0 Total ___/206. 221 | 2,983 |10,760 |------_ meet ee hy TAIL |e SHOS 4s Saeemes |Saeean| Sane Average} 7.365 OEE Teo Gish a) a ee (6653 | paaimenae 299), | eens eee | ee eee 8 This coal was unusually sooty depositing enough on the tubes in a few hours to burn off at this time. 332 COx. TABLE IV.—Observations in detail of the tests of coals—Continued. J.—FIRST TEST OF COAL FROM THE MILITARY RESERVATION, BATAN ISLAND— 100 FIRINGS DURING 73-HOUR TEST. [Test No. 10, Table II.] Average com- Kilos of ¢ Steam pressure position of flue pe Osea water fed to | Fire Baus: sone gases, in percent. boiler— taked Cjean- 5 2 d fire | ture of slie- {© Utes Aime alert = Aiie ed, Be Slane: ner Pounds) gases, | Dur- Dur- time stants per base of | ing ing after |“; Saaare square | stack COs | O2 | CO pee Total. pee Total. start- ing. TANG, inch. - | riod. riod. ing. Beginning__| 7.875 112 370 |) i hour _____- 7.242| 103 | 390 | a hour se 7.312 104 415 | hour 6.539 93 435 | hours sae 6. 890 98 430 | 1} hours __--_! 6.820 97 455 | 1; hours ____| 6.679 95 470 ee 0/ 3.4 1} hours _---_| 6.890 98 390 ) 2 hours =--__ 7.382 105 420 | | 2} hours ____| 7.593 108 460 | 2}hours__.| 7.804] 111 | 420 | 2? hours —---| 7.172 102 385 | 3hours__.| 7.101| 101 | 425 |J 3_hours ___- . 609 94 410 |) 70 865 275 3;hours___| 7,812] 104 | 495 | 65 | 930 | 275 3} hours __--| 8.156 116 420 | | 65 995 278 4 hours _____ 7.875 112 420 65 | 1,060 275 42 hours ____| 7.382 105 400 | | 65 1,125 275 4yhours_..| 7.593 108 | 420 || | 70 | 1,195 | 200 43 hours --_-| 7.453 106 400 | | 70 | 1,265 200 5 hours -_-- 7.312] 104 | 405 | 70 | 1,385 | 275 5} hours ----| 7.875 112 400 | 11.61 4.4] 0.0 } 65 1,400 275 5i hours-__| 7.875 | 112 405 | | | 65 | 1,465 | 275 | 5,600} 5 25 |______ 5# hours | 7.875 | 112 | 405 | | G5] 14580) |) 8275) | 5,8 7o)| = eee 6 hours ~___- 7.523 |. 107 400 | 65 | 1,595 Di || GTB) | a 6} hours ____| 7,382 105 415 | | | 65 1, 660 275 | 6,425 | 6 02 |-----_ 6; hours _-_-| 7.734 110 440 | | 65 | 1,725 275 7 6} hours ____} 7.523 107 425 | 65 1,790 275 7.875 | 112 410 | | 65 | 1,855 275 7-523 | 107 | 395 | | 65 |1,920 | 275 7. 734 | 110 365 L 27.8).1,947.8) 166 Total ---|229, 910 | 3,270 [12,825 |_------|___-_|___.- 1, 947, S|. ----- 7601S | toss al eae Rewenas Average| 7.417 105.5 A145 )> eee eee BEES 64:9| = et 296) aoe | Eee a PHILIPPINE COALS AS FUEL. 333 TaBLeE 1V.—Observations in detail of the tests of coals—Continued. K.—SECOND TEST OF COAL FROM THE MILITARY RESERVATION, BATAN ISLAND—64 FIRINGS DURING 7;s;-HOUR TEST. [Test No. 11, Table II.] fStiecien jae sine CeaoRtOL fade TROS Os Gon! Watenteclto Fire Baus e- Foie gases,in percent.) Purned— boiler— naked Clean: Time after quteiol slic: eae starting. aie Pounds fess, Dur- Dur- | time | eee square} snare |stack.| CO: | 2 | CO per | Total. | Fé | Total.| pe ing. meter. | 12ch. riod. riod. ing. | | | | Beginning__| 8.367 119 3 O91) teas | een | bases (de see Pee | ene et 2 hour______ 7.945 113 315 | 60 60 225 SOUL eees 7. 664 109 305 | 60 120 225 hour... 7.664} 109 | 305 || 60 | 180 | 225 1 hour _____- 7. 734 110 315 7.8 | 7.8 | 0.6 |{ 60 240 225 1} hours ____| 7.734 110 360 60 300 225 1: hours -___| 7.875 112 365 | 60 360 225 1} hours ____-| 7.875 112 340 60 420 225 2 noOuUTs = 7.593 108 320 J 60 480 225 2} hours --_-| 7.734 110 300 } 60 540 225 2% hours ___-}. 7.374 110 300 60 | 600 200 23 hours ____| 7.172 102 282 | 60 660 225 3 hours'==-== 7. 804 111 315 60 720 225 3} hours -___| 7.875 112 315 \ 7.5/7.3 | 0.9 59 775 200 33 hours -_--| 8.226 117 290 ' 55 830 200 3} hours -__-| 7.523 107 300 55 885 200 4 hours _____ 7.945 113 290 ie) 940 200 41 hours --__| 7. 664 109 310 55 995 200 42 hours ----| 7.875 | 112 320 (55 | 1,050) 200 4} hours -_--| 7.382 105 288 5d 1,105 | 200 5 hours ____- 7, 945 113 310 Bb) 1,160 | 200 51 hours -___| 7.664 109 300 55 1,215 | 200 54 hours -__-| 8.015 114 310 Bd) 1,270 | 200 5} hours --__| 7.734 110 288 55 1,325 | 200 6 hours _____ 7. 523 107 310 822) 1)65/2) | 1) 55 1,380 | 200 6} hours —---| 7.804 111 300 55 1,485 | 200 6; hours ----| 7.734 110 300 55 1,490 | 200 6} hours -_-_| 7.523} 107 | 310 | 55 | 1,545 | 200 TAD OUTS ee 7. 734 110 310 | 42.6) 1,587.6) 158 ds OUTS == | Sass oes |en a eees [eee 0 | 1,587.6 0 Motall=|225406 10 |t3¥ 2011184983 || mene | eee eee S70 ene 518888 |useeaes bees aee Average! 7.761 110545310) ees ee On| aaa 20833) isc 2 ei 75034——3 334 COxe TaBLe 1V.—Observations in detail of the tests of coals—Continued. L.—THIRD TEST OF COAL FROM THE MILITARY RESERVATION, BATAN ISLAND— 56 FIRINGS DURING 7-HOUR TEST. [Test No. 12, Table II.] steampresmre| | avergecam. | xuosorcoat | yA, | ine gauge. Tem-|sases, in percent, ured boiler— raked Gjoan- ) re ee ie Time after Pe flue | al time starting. ae Pounds) gases, Dur- Dur- time | etter) square anane Raselof CO. | O2 | CO ae Total. ae | Total. etter ing. ces inch. riod. riod. ing. | Beginning__| 7.804 111 300 (fee ae ee eee + hour -_-_-_ 7. 664 109 330 60 60 225 ehOUr eee 7.7384 110 330 60 120 225 3} hour_____- 7.945 | 113 | 345 || 60 | 180 | 225 1shoursa= 7,945 113 355 60 240 225 13 hours ____] 7.875 112 330 60 300 225 | 14 hours ____| 7.664 109 315 60 360 225 1} hours ____; 7.875 112 320 |$ 5.0] 8.0] 1.4 60 420 225 | 2 hours ____- 7.804| 111 | 290 60 | 480 | 225 | 21 hours ____| §.015 114 285 | 60 540 225 23 hours ____| 7.312 104 435, 60 600 225 23 hours ____| 8.015 114 380 55 655 225 3 hours _____ 7.875 112 345 55 710 225 32 hours ____| 8. 086 115 320 55 765 225 31 hours ____| 7.875 112 350 55 820 225 33 hours -___| 7.945 113 310 55 875 225 4 hours _____ 7.382 105 290 53 930 200 4} hours ____| 8.297 118 350 55 985 200 4} hours ____| 7.312 104 340 55 | 1,040 200 4} hours ____}. 7.875 112 315 55 | 1,095 200 8.015 114 330 55 =| 1,150 200 52 hours ____| 7.523 107 300 55 | 1,205 200 53 hours ____| 7. 664 109 325 8.8 | 6.8} 0.4 }{ 55 | 1,260 200 5}? hours ____| 7.734 110 330 55 1,315 200 6 hours _____ 7.523 107 390 5d | 1,370 200 6: hours ____| 7. 664 109 340 55 | 1,425 200 63 hours ____| 7.593 108 335 j 55 1,470 200 6} hours ____| 7.664 109 365 55 | 1,535 200 7 hours —-_=- 7. 382 105 340 J 27.6) 1,562.6; 178 Motal2s|225x0G15 835,201 893690) a ee |e 1, 562. 6 ae At yen Peer Meee one (ee nee | Average! 7.761 a1 OFA 3Bay| le eee | Seceie Bee ae 5018 eens PAB) eee eS ee PHILIPPINE COALS AS FUEL. TaBLE IV.—Observations in detail of the tests of coals—Continued. 335 M.—FIRST TEST OF LUMP COAL FROM THE MILITARY RESERVATION, BATAN ISLAND—76 FIRINGS DURING 7-HOUR TEST. [Test No. 13, Table II.] Steam pressure cette Kilos of coal | eee to | Fire gauge. Tem- |gases, in per cent. louie | boiler— raked) Gy ean- ener | or led fire Time after = ae or | eu time’ SLU nie Pounds) gases, Dur- Dur- time atten square Renae paseof C02 | O2 | CO ne Total. oe Total. eter ing. cen ue inch. riod riod. ing. h,m. Beginning__| 7.172 102 0 00 + hour -__-__ 7. 664 109 65 65 4 hour _____- 7. 664 109 60 135 }# hour --____ 7.593 | 108 395 60 195 1 hour ___._- 7.528 | 107 | 430 60 | 255 1+ hours --__| 7.804 111 440 11.6 | 6.2 | 0.6 14 60 315 1} hours ----| 7.523 107 450 60 375 1} hours --_-| 7.382 105 445 |! es) 430 2 hours —____ 7. 784 110 448 55 485 2} hours 7.664 | 109 | 397 55 | 540 24 hours —-_-| 7.453 106 430 | 60 600 23 hours _---| 7.242 103 442 60 660 3 hours __---| 7.523 107 465 55 715 32 hours -___| 7.382 105 440 5d 770 3% hours ----| 7.523 107 435 55 825 3p hours -__| 7.734| 110 | 420 |f27°| 7?) %419 55 | sao 4 hours _____| 8.226 117 405 55 935 42 hours ___-| 7.875 112 378 55 990 4 hours ____| 8.307 118 385 55 | 1,045 ~ 43 hours ____| 7.382 105 370 |) ( 55 | 1,100 5 hours _____ 7.945 113 385 55 | 1,155 5} hours _-_-| 7.804 111 400 50 | 1,205 53 hours --_-| 7.312 104 400 60 | 1,265 5} hours --_-| 7.523 107 380 5 60 | 1,325 6 hours _____ 8.507 121 390 S| to ORE 55 | 1,380 63 hours ____! 8.015, 114 400 5d} 1,435 6} hours ___-| 7.664 109 412 55 | 1,490 6} hours -___| 7.664 109 415 55 | 1,545 7 hours —__-- 7, 242, 108 432 27.5) 1,572.5 Total = =|220: 046) ))3; 158) |i, 194) |e ee |e es S725 | eee Average) 7.656 1O8%9 | RS 414 eos | ee ee 0s) aaa 336 COX. TABLE I1V.—Observations in detail of the tests of coals—Continued. N.—SECOND TEST OF LUMP COAL FROM THE MILITARY RESERVATION, BATAN ISLAND—89 FIRINGS DURING 6§-HOUR TEST. - [Test No. 14, Table II.] Steam pressure Eee Kilos of coal “<4 algsioh: o | Fire gauge. Tem- |gases,inpercent.) PUrmed— boiler— TAKed G)egn ees pe, Aft, edie, Time after | fie | ed, | time SUTIN oe Pounds! gases, | Dur. wae | | time | Biter square square base ot co. | 0, | CO ae Total. pe eo age ing. Peter inch. | | riod. riod. 3 | Beginning-_| 8.156| 116 | 400 |} | cnc Sea ree | ves 2 hour __---- 7.734 | 110 | 480 | || 60 | 60 | 300 hour ___-_- 8.086 | 115 | 430 || 55 | 115 | 300) 3 hour_____- 8.086 | 115 | 430 52 | 167 | 300 | 1 hour ------ 7.664 | 109 | 388 52 | 219 | 300 | 1:hours__| 7.664 109 | 392 | 52 | 271 | 300) 1} hours --_-| 7.664 109 408 | }10.0| 7.4 | 0.2 52 323 300 | 12hours_---| 8.437| 120 | 426 | 52 | 375 | 300 | 2 hours ____- 7.804 | 111 | 393 52 | 427 | 300 21 hours____| 7.945 | 113 | 380. | 52 | 479 | 300 91 hours ----| 7.382] 105 | 375 52 | 531 | 300 23 hours ----| 7.593 | 108 | 410 | 52 | 583 | 300 3 hours ____ 7.664 | 109 | 400 |J 52 | 635 | 300 Bt hours....| 8.015] 114 | 370 |) f 52 | 687 | 300 32 hours_.| 7.804) a | — 989 | 52 | 739 | 300 33 hours____| 8.015 114 400 52 791 300 4 hours —___- 8.086} 115 | 380 | 52 | 843 | 300 42 hours ____| 8.156 116 344 52 895 300 43 hours____| 7.875 | 12 | 402 | 52 | 947 | 300 | 43 hours____| 8.086] 115 | 380 52 | 999 300 5 hours ____- 8.015| 14 | 386 |} 8.0 110.6) 0.0\4 52 f,o51 | 300 52 hours ____| 7.664 109 372 52 |1,103 | 300 5 hours___] 7.382| 105 | 390 52 [1,155 | 300 52 hours --_-| 7.523] 107 | 380 | | 52 [1,207 | 300 6 hours _____ 8.367| 119 | 384 52 [1,259 | 300 6 hours ___| 7.664] 109 | 397 | 52 1,311 | 300 6 hours | 8.015 | 114 | 385 | 52 (1,363 | 300 | 6% hours ____| 7.734} 110 | 365 ( 41.311,404.3) 185 GB. rours Meee e7 e734 | fect | een oe | nee net en 0 [1,404.3] 0 Total ___|228, 014) 3,243. [10,9860 |_______|_____|_____ 1, 404, 3] 7ige5 || see ese! (iecurae Average| 7.862] 111.8) 392,4|______|____|_____ 5isd|eaeet 252) een Eases: Jone i | a The fire on the grate was not disturbed during the entire run. PHILIPPINE COALS AS FUEL. 337 TaBLe 1V.—Observations in detail of the tests of coals—Continued. O.—FIRST TEST OF COAL FROM BETTS’ MINE, BATAN ISLAND—66 FIRINGS DURING 4-HOUR TEST. [Test No. 15, Table II.] | x Average com- - Klos of ce Seo ee y position of flue xalosioticoal water fed to | Fire | gauge. ee gases, in percent. boiler— raked! Glean- a 5 d fire ae ture of b slic- |E@ Ure, nes Kilos alee Gel, aiter s- per Pounds gases, | Dur- Dur- time stant square eae ase foi CO, | O. | CO ae Total. ae Total. ae ing. cone inch. riod. riod. ing. | Beginning_-} 6.187 88 440 |) (eRe oe et Lhour------| 5.765| 82 | 465 | | tB 75 | 275 1 hour --_--- | 5.765] 82 | 412)\11.4| 7.4] 0.0|4 75 150 | 275 hour _----- 5.273] 75 | 403 | 7 295 | 275 Lhour ___--- 5.203] 74 | 540]) 7 300 | 275 1} hours ---_-} 5.273 75 520 } 75 375 | 275 1 hours__..| 5.062] 72 | 480 ley slaolaoll 2 450 | 275 13 hours ___-| 4.851 69 450_ pace Mae 75 525 | 275 2 hours --___ 4.008 57 430 75 600 | 275 2} hours ._--|_ 3.937] 56 | 425 tB 675 | 275 21 hours_-_.-| 4.078| 58 | 415 iB 750 | 275 23hours----| 4.008] 57 | 405 jes Gee bet yr 820 | 275 “| 3.867 | 55 | 360 70 890 | 275 3} hours | 3.656} 52 | 363 |, 70 960 | 275 3h hours-.-| 3.164] 45 | 292]| — 70 | 1,030| 275 33 hours --| 2.953} 42 | 250 i 7-8/10:0'| 0-8) 79 | 4,100) 975 4 hours —_-_- 2.953 | 42 | 250 |) 34 | 1,134] 186 Total ___ Average 338 COX. TABLE I1V.—Observations in detail of the tests of coals—Continued. P.—SECOND TEST OF COAL FROM BETTS’ MINE, BATAN ISLAND—125 FIRINGS DURING 7-HOUR TEST. [Test No. 16, Table II.] steampresore| | ANGRECSOR, xcosof coal | AMORA, | wire Les Tem- |gases, in per cent. igure boiler— —.Taked) Gyog yn Time after thre of ] ] slic: ns { starting. ee Pounds ee Dur- | Dur. fae ater anare square eee or COs |. 0; CO pes Total. | pe Total. ae ing. meters inch. riod. riod. ing. | | Beginning__| 7.382 105 430 |) 2 hour _____- 7.875 | 112 | 455 | 2 houre=2==5 6. 961 99 455 3 hour ___-- 7.453} 106 | 480 || 1shoureesss= 7.031 100 460 | -10.9 | 6.9 | 0.0 1 hours ----| 7.172} 102 | 450 | 13 hours ----| 6.539 93, 440 1} hours _--_| 6.609 94 415 | 2 hours —___- 7,031 100 440 | J 21 hours -_-_| 7.593 108 420 23 hours ----| 7.172 102 430 23 hours --_-| 7.172 102 425 3 hours ----- 7.382 105 450 32 hours -__-| °7. 453 106 430 | 3: hours __--| 7.664 109 440 110. 6 | 7.4 | 0.6 3? hours ____|_ 7.593 | 108 445 | | 4 hours -__-- 7. 804 111 440 42 hours ----| 7.523 107 400 | | 41 hours ____| “7.242 103 365 4} hours ----| 7. 664 109 430 5 hours —---- 7. 734 110 405 5} hours ----| 7.523 107 435 53 hours ---_| 7.453 106 430 5} hours -_--| 7.945. 113 435 6 hours ____- 7.453 | 106 | 450/ $11.0 | 7.4 | 0.0 63 hours --__| 7.664 109 430 63 hours ____] 7.382 105 480 63 hours ____| 7.523 107 490 | 7 hours _._- | 6.890 98 505 | J Total -__/213. 882 | 3,042. 12,760 |_______ [S34 ee Average| 7.375] 104.9] 440 |____-_- |ece| ee | | PHILIPPINE COALS AS FUEL. 339 TABLE IV.—Observations in detail of the tests of coals—Continued. Q.—FIRST TEST OF COAL FROM THE COMANSI MINE, NEAR DANAO, CEBU— 61 FIRINGS DURING 54-HOUR TEST. [Test No. 17, Table II.] Steam pressure Soe orate Kilos of coal ORF Fire 5 gauge. Tem- |gases, in percent.) PUTed— boiler— raked) (yaa. | ee L ff, edie, Time aiter 4 fle al time starting. ee Pounds| gases, Dur- Dur- time ered square a quae pasctor CO, | O, | CO Be Total. pe Total. stat ing. THER, inch. riod. riod. | ing. Beginning__| 6.961 99 325 57 } hour —--__- 7. 593 108 385 57 + hour ______ 8.015 114 400 57 # hour/==-2— 7.453 106 385 E 57 phous 7.312 | 104 400 |} 9.0] 6.5] 2.0 |4 57 1 hours --__} 7.598 108 495 57 14 hours ____| 7.242 103 430 57 1} hours ____} 7.528 107 390 57 2 hours —--__ 7.312 104 385 57 2} hours ____| 7.804 111 400 |) | f 57 2} hours -___| 7.664 | 109 455 || 57 23 hours --__| 7.593 108 375 57 3 hours ----_| 7. 453 106 390 57 ahours._.| 7.172| 102 | 330 |{-"°| “°| 4 | 57 3; hours ----] 8.086 115 390 57 3} hours --__| 8.015 114 450 57 4 hours _____ 7.382 | 105 380 |J 57 4} hours ____| 7.804 111 360 57 4% hours --_-| 7.453 106 370 57 43 hours --__| 7, 453 106 379 57 6 hours _____ 7. 664 109 350 ' HAG Bebe? ca) 57 52 hours --__| 7.453 106 375 57 1,197 Ee, OS0 pee sce 54 hours --__| 7.593 108 375 30.4) 1,227.4) 146.66, 236. 6)-_____ 5 30 Total _-_|173. 593 | 2,469 | 8,970 |_______ eel aera 221 xe | eee GS 236%6 | seem eet ims Average] 7.547 107.3 D040 | see 28354| Sos pees a - 340 COX. Taste 1V.—Observations in detail of the tests of coals—Continued. R.—SECOND TEST OF COAL FROM THE COMANSI MINE, NEAR DANAO, CEBU— 68 FIRINGS DURING 74-HOUR TEST. [Test No. 18, Table II.] Steam pressure See orate Kilos of coal | ,, log ot to | Fire gauge. ae gases, in per cent. burned boiler— eked! Clean- Time after puretot au leery starting. a Pounds] gases, Dur- Dur- time, pee pours Eonar Pager co, | O, | CO pe | Total. Be Total. start ing nto inch. riod. riod. | ing. | Beginning__| 7.523 107 325 |) | + hour -----_ 7.875 112 355 | + hour ___-_- 8. 086 115 345 | j} hour _____- 7.945 | 118 | 330 | hhounsse= 7.453 106 330 1t hours _._| 7.875| 112 | 360 | 13 hours ----| 7.593 108 360 1 hours____| 8.015 | - 114 S30 /i|p seen bas ae | (a) { 55 2 hours ____- 7.734 | 110 | 355 | 52 2% hours -_-_| 7.453 106 330 | | 52 2} hours --_-) 8.015 |, 114 315 52 23 hours ___-| 7.593 108 325 | 52 3 hours _____ 7.593 108 330 52 32 hours -_-_| 7.382| 105 | 315 | | 52 3; hours __-_| 7,945 113 330 l 52 3} hours —-_-| 7.453 106 330 ( 52 4 hours _____ 7.875 112 495 52 4+ hours ----| 7.945 118 415 52 4: hours ----| 7.875 112 325 52 4} hours ___-| 8.367 119 350 52 5 hours ~--__ 7.945 | 113 320 | | 52 52 hours -__-] 7.523 107 315 | 52 5} hours _-..- 7.453 106 320 52 5@hours___| 7.664 | 109 | 315 | {-~~-[>-— |} 59 6 hours —___- 7. 382 105 310 | 52 | 6 hours ____| 7.664] 109 345 | 52 6} hours ----| 7.523 | 107 330 | 52 6} hours ----| 7.664 109 350 | 52 7 hours - 7.593 108 365 52 73 hours --_-] 7.382 105 365 | | 52 73 hours ----| 7.523 | 107 327 97 Total ___/238.911 | 3,398 10,612 |---_---|=----|___. ~|1, 566 Average| 7.706 LOSTG|S 6.342) |Baweees ees P= 52.2 PHILIPPINE COALS AS FUEL. 341 DISCUSSION. The data sustain the conclusions that the value of a coal for producing steam in an ordinary boiler is determined not only by its fuel ratio and by the total number of heat units set free during its complete combus- tion, but it is also dependent largely upon other and variable factors. Impurities in the coal—Vhe purity of the coal—that is, the admixture of earthy matter, moisture and other foreign material which it contains— is an important consideration. If the percentage of ash and water is small the theoretical heat value of the coal is proportionally increased and from a commercial standpoint the original cost of freight and han- dling per thermal unit and the expense of removing the ash as well is correspondingly decreased. ‘These items represent a direct saving. More- over, with coals high in moisture the efficiency is lowered directly by the specific heat of the water. The color of the ash indicates the iron content and is also usually taken as an indication as to whether or not the coal will clinker. How- ever, iron is but one constituent and other factors enter in just as they influence the fusion point of clay** or cement. As comparatively few coals burn without forming clinker, it is interesting to note that in many of the tests of Philippine coal, in particular the tests of the coal from the military reservation, Batan Island, where the percentage of ash is high and it is brick-red, very little clinker was produced. It is probable that the ash bed in this non-coking, highly volatile coal is not heated sufficiently high to form clinker. he distillation of volatile matter is endothermic and therefore the explanation of the lack of clinker is probably partly to be found in the fact that the distillation of this large percentage of volatile matter keeps the temperature of the fuel bed low. Furthermore, in a non-coking coal the lumps are thoroughly disintegrated with the expulsion of the volatile matter and the ash kept cool by the air and gases passing through and around its particles. If the same ash were in a coking coal it would be held in the lump and probably be heated hot enough on the grate and in the fuel bed to melt it and produce clinker. It is believed that a reasonable amount of ash has little influence on efficiency other than the amount of combustible carried away, except where it interferes mechanically. If a coal clinkers and tends to close the air spaces it greatly increases the labor in connection with its con- sumption and entails a loss of heat through the furnace doors through frequent opening to work the fires. On the other hand, although clinker “Cox, A. J.: The occurrence, composition and radioactivity of the clays from Luzon, P. I., This Journal, Sec, Ae (1907), 2, 427. 342 COX. may hinder combustion, it prevents fine coal from falling through the grate and in this way may partially compensate for its inconvenience. The finer and dirtier coal from Batan Island after correcting for loss of fine coal (i.e., calculated to coal actually burned), and the difference in ash content, gave somewhat lower efficiencies than the larger and care- fully selected sizes. The only apparent difference in the behavior and quality of the various sizes is that the fine coal, high in ash, tends slightly to smother the fire and steam can not be produced at as great a rate as with the larger sizes. An inspection of Table II shows that the first test of the coal from the military reservation with the highest percentage of ash has a less evaporation per unit of combustible actually consumed than the second and third, which contain less ash, and still less than the fourth and fifth which contain still less ash. The variation, however, is not believed to be due to the ash, but is largely accounted for far more easily fixed carbon ’ volatile combustible matter the greater ratio giving the greater efficiency; although that very high ash may reduce the draft, cause a slower rate of combustion and there- fore less complete combustion in the furnace chamber and the range of the water tubes is not without reason. Fire box and grate-—TVhis Bureau has what is ordinarily considered to be a good boiler plant. However, it has a short fire box and only the usual vertical baffling and this is not sufficient to enable it to be run with- out some black smoke and loss. It is a recognized fact that the loss of heat due to the actual carbon in the escaping gases is small, perhaps never more than 1 per cent, but smoke is a strong indication of the presence of combustible gases the loss of which may amount to several per cent and materially impair the efficiency. A short fire box is not at all suited successfully to burn Philippine coal. I have often urged 1* the necessity of a setting with an elongated fire box and combustion chamber for burning this class of coal. The combustion space must be long and large enough for the combustible gases and air to mix thoroughly and to produce complete combustion. The United States Geological Survey has expressed the same opinion and further lays special emphasis on the necessity of an additional baffle wall.t* Such a wall would undoubtedly cause more perfect mixing and therefore more perfect combustion, which is the desired end. It is probable that eddies such as one seeks to attain in a reverberatory furnace, caused by any obstacle in the path of the gases, greatly aid the mixing. Any scheme which works in the direction of retarding the by a consideration of the fuel ratio, i. e. * Cox, A. J.: This Journal (1906), 1, 877; Sec. A. (1907), 2, 41. *U. 8. G. S. Bull. (1907), 325, 62. PHILIPPINE COALS AS FUEL. 343 exit of the gases of the flame stream until combustion of the volatile combustible matter is completed in the combustion chamber, contains the possibility of greatly increasing the efficiency of Philippine coals. Satisfactory baftle walls would probably be of as much value as a con- siderable increase in the length of the fire box. A boiler with the same setting as those of this Bureau, but arranged with different baffling forming a tile-roof furnace, has been used on Illinois coals and is said to run at capacities of from 50 to 100 per cent without smoke." Various grates other than the ordinary bar have been suggested and tried on coals of the sub-bituminous variety. It was hoped that the per- forated grate would be more economical of coal. However, in the tests of Mr. Betts’ coal there was a slight incipient clinker which could not be dislodged from the holes and the steam pressure fell at the end of the test because of lack of draft. It was not possible to experiment much with this coal beforehand and but little information regarding it could be obtained. The grate worked well with Australian coal. With more experience and slight modifications this may still be more satis- factory than the ordinary grates. Mr. Betts has tried a herring-bone grate which he reports to be very successful. The advantage of a grate of this type over the ordinary gridiron is that shorter, thinner and more bars may be used without danger of their melting down and in this way the air spaces increased in number, but diminished in size without changing the ratio between air space and grate surface. It has also been suggested that the loss of combustible matter in the ash could be prevented by burning these coals on a rocking grate. It is hoped that the study of the behavior of Philippine coal and coals of this class will soon result in the discovery of a more satisfactory grate and a method of combustion that will be more economical of the coal. Reconstruction of the present boiler settings in the Archipelago is out of the question. Greater efficiency, therefore, can be obtained only by building additional baffle walls, usmg a more satisfactory grate, elongating the fire box or heating the air before entering the grate, and these improvements from an economic standpoint can best be tried in the order of enumeration. * Breckenridge, JL. P.: Univ. of Ill. Bull. (1906), 4, No. 31, 22. M. Ernest Schmidt, Bull. soc. ind. @Amiens, 2-3, 102; C. A. (1908), 2, 174, has called attention to the fact that it is diflicult to destroy smoke after it is once formed, but believes in preventing its formation by gradual introduction of coal into the fire box, if possible under the burning combustible, and finally, by the use of a mass of fire brick kept at a high temperature. He also considers the heating of the air before entering the grate necessary. In the combustion of Philippine coal where high chimney temperatures are obtained this might be accomplished by a down-draft pipe through the stack. 3 344 COX. Fic. 2.—IpEAL SECTION SHOWING ADDITIONAL BAFFLE WALL AND AN ELONGATED Frre Box. In the plant of this Bureau, Australian coal burns to a large extent on the grate, while most of the Philippine non-coking coals containing high volatile matter are at a disadvantage, as they burn to a very much greater extent in the combustion chamber. An inspection of the fore- going tests of the coals from Australia (Westwaldsend), Batan Island (Military Reservation and Betts’), and Cebu (Comansi) will show that our boiler-plant is unfavorable to Philippine coal. This may the more readily be seen from the following table: TABLE V. Equivalent evaporation of Calorifie |Equiyalent| water from value of evapora- | and at 100°C. the com- tion of per kilo of bustible in| waterfrom| combustible calories and at actually con- Source. as deter- | 100°C. per| sumed, anti- mined in a kilo of cipated from Berthelot-| combus- | the calorific Mahler tible value when bomb ealo-| actually Australian rimeter. |consumed.| coal is taken as the base of comparison. Australian: (Westwaldsend); average of tests 1, 2, and 3; Table Tl 22-322 eee 7,791 8, 688 8, 688 Batan Island: Military reservation; average of tests 10, 11, 12, 13, and14, ‘Tables. = Sse ee 7, 166 6,773 8, 000 Betts’; average of tests 15 and 16, Table II___--_-___ 6, 297 6, 698 7, 020 Cebu (Comansi); average of tests 17 and 18, Table II____ 7, 207 7,122 | 8, 040 POLLO FS teste ye Ta] ey ee en pea h}0| a 8, 210 PHILIPPINE COALS AS FUEL. 345 Coals which burn low and close to the grate give greatest efficiencies ; those which burn high lose much through the grate, give low initial temperature in the fire box, leaving the fuel bed comparatively cool, and the result is combustion at the rear of the chamber, imperfect heat absorption and therefore-low efficiency. I think this loss is largely due to the type of boiler, and one should be constructed for these coals that would obviate these losses. I should like to be in a position absolutely to name the best class of furnace for each coal, but not enough tests have been carried on to enable me to do so; however, considerable information as to the best form of furnace has been given. Loss through the grate-—A portion of the combustible matter of the coal falls through the grate into the ash pit and is not burned. For a definite coal this varies with the grate and for a certain grate it varies with the coal. It is a most difficult task, not yet accomplished, to con- struct a grate that is suitable under any and all conditions of operation. Owing to my inability to have a grate suitable for each coal this discrepancy is much larger in some cases than in others, and therefore I have given, in addition to the usual data, recalculated results to show- the values when this factor is eliminated, i.e., as if this amount of coal had never been fired. Draft, chimney gases and loss through the stack.—Draft, measured by the reduction of pressure as compared with that of the atmosphere, which depends on the relation of boiler, furnace, grate and stack, largely controls the air which enters and the value of the fuel is influenced by it to a marked extent. However, in a boiler plant in the tropics much depends on the direction of the wind, since in most cases the boiler is not protected at the sides. Too much air is better than too little; on the other hand, an excessive amount dilutes the gases, lowers their temperature and increases the waste to the stack by an amount equal to the specific heat of the moisture from the excess of air and the heat carried away by the additional quantity of dry chimney gases. The loss up the chimney decreases and the efficiency rises with a reduction in the supply of air until a point is reached at which the loss due to slightly incomplete combustion is just equal to the gain obtained by decreased loss to the stack. Beyond this point the decrease in efficiency is very rapid. It has been my aim to regulate the air supply as much as possible without reducing the completeness of combustion, and in that way I endeavored to control the quantity of gases leaying the system and therefore the waste heat. Without experience with a given coal it is not always possible accurately to supply the proper amount of air for its ideal combustion. It may be noticed from an examination of the tests that a certain amount of carbon monoxide was observed in the chimney gases. This amount was greatest in those from the coal from _ 346 COX. the Comansi mine at Danao, Cebu (test 17) where there was an abnormal waste to the stack and the efficiency recorded is therefore probably somewhat low. ; It has been shown*® that any considerable percentage of carbon mon- oxide is threatening to efficiency. Owing to the infiltration of an unknown quantity of air no exact limit could be set to this, but since the presence of carbon monoxide may also be taken as an indication of other incomplete combustion losses, high carbon monoxide is a prominent danger signal. It has also been shown™ that the furnace efficiency drops very rapidly after the carbon dioxide content in the flue gases has reached about 9 per cent or perhaps 12 per cent if the gas has not been diluted by leaks. From a knowledge of the law of mass action one would expect, where the oxygen content is low and the carbon dioxide high, that some carbon would only be partially oxidized, that is, the presence of some carbon monoxide would be probable; however, an equi- librium may not always be attained in the combustion chamber. As the flue gases passed the sampler in the seventeenth test the oxygen content -was higher and carbon dioxide lower than in the tenth where combustion was complete. Such a condition as that in the seventeenth, where the gas analyses represent the average of a period, might be produced by careless stoking so spasmodic that at times the percentage of oxygen would be small, with incomplete combustion, and at other times so large, that the average oxygen content would be increased. However, I do not believe that this is the case in this series. An explanation which suggests itself is that each individual coal, at any given temperature, may require a certain excess of oxygen, varying with the complexity of the hydrocarbon compounds, to effect complete decomposition of the coal gases. If the latter pass the high temperature of the furnace undecomposed, then the small supply of oxygen is not sufficient to effect combustion before they escape from the combustion chamber. Furthermore, owing to the coolness of the fuel bed and combustion chamber when highly volatile coals are burned, combustion takes place slowly and it is not surprising that the carbon monoxide and other combustible gases are swept on and cooled below their ignition tempera- tures before combustion is complete. The corrected ignition temperatures of various molecular relations of hydrogen and carbon monoxide, with oxygen are the following :** 4H,+0,=605° 6CO+0,=721° 2H,+0,=540° 4C0+0,=628° H,+0,=514° 2C0+0,=601° H,-+20,=530° CO+0,=631° H,+40,=571° *U.S. G. S. Bull. (1907), 325, 65. THbids ole 8K. G. Falk, Ann. d. Phys. (1907) (4), 24, 450. PHILIPPINE GOALS AS FUEL. 347 The introduction of an inert gas such as the nitrogen content of the combus- tion chamber, greatly raises the ignition temperature and for the bimolecular reaction between hydrogen and oxygen it is increased according to the equation T=T’+30 n where volume of the nitrogen (N,) ™— volume of the hydrogen (H.) or the oxygen (O,) whichever is present in the smallest quantity. For the trimolecular reaction between carbon monoxide and oxygen the ignition temperature is increased ,according to the equation T='T’+80 n’ where volume of the nitrogen (N,) ™ volume of the carbon monoxide (CO). , The temperature coefficient of the reaction velocities for an increase of 10° is 1.31 between the limits 514° and 550° for a mixture of hydrogen and oxygen; and 1.24 between the limits 601° and 645° for a mixture of carbon monoxide and oxygen. The introduction of an indifferent gas (nitrogen) reduces the magnitude of this coefficient in proportion to the quantity added. For a mixture of two volumes of carbon monoxide and one yolume of oxygen Helier” gives the following maximum formation of carbon dioxide, expresed in per cent at the given temperature: Degrees | Per cent Degrees | Per cent centigrade.| CO» centigrade.| CO»s | 195 0.13 504 | 7.8 302 0.44 || 566 | 14.43 365 1.41 575 17. 27 408 3.03 600 21.14 ! 418 3.41 689 43. 36 | 468 4. 64 | 788 60.3 | 500 6.2 || 855 65.0 | The formation of carbon dioxide from the carbon compounds in coal or eyen by burning carbon monoxide itself is no simple one. The dissociation of carbon dioxide into carbon monoxide and oxygen and the part that water plays in the reaction must all be considered. A perfectly dry mixture of carbon monoxide and oxygen can neither be exploded by means of a red glowing platinum spiral nor an induction spark.” The particles of water themselves play an important part in the reaction. Even at ordinary temperatures there is a small amount of free hydrogen and free oxygen in water vapor. The equilibrium at 10° contains one volume of free hydrogen and one-half volume of free oxygen for every 4.55. 10° volumes of water vapor. The higher the temperature the greater the amount of uncombined gases in proportion to water vapor. When the equilibrium is reached at 100° there is one volume of free hydrogen and one- half volume of free oxygen for each 1.14.10" volumes of undissociated water vapor. At very high temperatures free hydrogen and oxygen are present in such quantities that they may be directly determined. These free gases are chemically very much more active than the water molecules themselves. The “Ann. de Chim. (1897) (7), 10, 521; Chem. Centrbl. (1897) I, 68, 487. 2? Dixon, Chem. News (1882), 46, 151. *t Bodliinder: Ahren’s Samm. chem. u. chem, tech. Vortriige (1899), 3, 388. 348 COX. oxygen unites readily with carbon monoxide to form carbon dioxide or the hydrogen with oxygen to form water or hydrogen peroxide. If the dissociation equilibrium is disturbed in either of these ways, more water molecules dis- sociate into hydrogen and oxygen atoms. When a temperature of the furnace is reached where this dissociation takes place faster than the dissociation of the oxygen molecules of the air, we have an explanation of the catalytic action of water in the combustion of coal and why a high combustion chamber tem- perature is desirable. In the combustion of a highly bituminuous coal, the extent of the loss due to the carbon monoxide and hydrocarbon gases of the gasified coal, passing up the stack before combustion is complete may be seen by an examination of the following table: Element. Product of combustion. | combustion in calories.22 | Heat of | | 1 Carbonies see Carbon monoxide -_______ | 2,435 | Doss sate Carbon dioxide —___---__- 8, 140 Eby nor ena eas aan Water 2-222 ees 34, 180 It will be observed that each unit of carbon burned only to carbon monoxide will result in a loss of 5,715 calories (over half) and each unit of hydrogen unburned will result in a loss of 34,180 calories. In these experiments this loss has been regulated as well as possible with the dampers and air supply at my disposition, but a difference in construction of the boiler plant would seem advisable for some of the varieties of coal. Approximately perfect combustion can be obtained by proper boiler and furnace design, construction and operation. An extremely rapid rate of evaporation, a low chimney temperature and completeness of combustion are incompatible. Messrs. Brecken- ridge, Parr and Dirks** found that the maximum rate of evaporation was obtained with the boiler running at its rated capacity, with the flue-gas temperature at about 260° C. With an increase in the rate of combustion the flue-gas temperature increased and the evaporation dropped off. Most of the Philippine coals easily gave a rate of evapora- lion equal to that obtained with Australian coal on an ordimary run. Absorption.—Highly bituminous coals are likely to cause a deposit of soot which reduces the efficiency of the heating surface. Boilers must be thoroughly cleaned before beginning tests. The necessity for this precaution is evident in that if the drum and tubes are insulated from the hot gases on the one side by a layer of soot and from the water on the other by a layer of scale, the absorption will be imperfect and the greater this insulation the more resistance to absorption and the greater # Calculated from the numbers of J. Thomsen: Thermo-chemische Untersuch- ungen (1882), 2, 52, 283 and 288. SUniv. of Ill. Bull. (1906), 3, 39. PHILIPPINE COALS AS FUEL. 349 the loss to the stack by the gases escaping at too high a temperature as compared with that of the steam in the boiler. Breckenridge et al** from results of boiler trials made to determine the effect of soot deposits on the evaporation in a horizontal tubular boiler conclude that it is not very marked. They found that the soot burned upon reaching a certain thickness, leaving but a very thin layer. Eyen with frequent and perfect sweeping of the tubes, no boiler cools the furnace gases to the temperature of the steam, but a certain amount of this heat waste may be recovered and the efliciency somewhat raised by the use of an economizer in the stack. The effect of scale on the transmission of heat through boiler tubes is very variable, the mechanical structure of the scale being at least as important a factor as the mere thickness. Schmidt and Snodgrass** have investigated this effect on locomotive boiler tubes and feel warranted in summing up the results of their tests in the following conclusions: “]. Considering scale of ordinary thickness, say of thicknesses varying up to one-eighth inch, the loss in heat transmission due to scale may vary in individual cases from insignificant amounts to as much as 10 or 12 per cent. “2. The loss increases somewhat with the thickness of the scale. “3. The mechanical structure of the scale is of as much or more importance than the thickness in producing this loss. “4, Chemical composition, except in so far as it affects the structure of the scale, has no direct influence on its heat transmitting qualities.” Boiler pressure.—The true boiler efficiency is the ratio of the heat absorbed to the heat which is available to the boiler; that is, that portion of the heat in the furnace gases which is above the temperature of the steam. From this it is evident that the higher the working pressure—that is, the higher the steam temperature—the less difference between a fixed temperature of the furnace gas and that of the steam and therefore the less heat available to the boiler. In order to obviate this difference in efficiency I have tried to maintain approximately the same steam pressure in the various tests. In those cases where there is a deviation, the efficiency attained is greater or less than the average accordingly as the steam temperature is greater or less. The facts have not been established giving the exact value of the effect for all changes in steam pressure upon the evaporative efficiency of a boiler. Goss*° has shown that “changes in steam pressure between the limits 120 pounds and 240 pounds will produce an effect upon the efficiency of the boiler which will be less than 0.5 pounds of water per pound of coal.” The difference is not large for the small ranges of pressure common in stationary practice; and although slightly more heat is available and 74 Loc. cit. > Univ. of Ill. Bull. (1907), 4, No. 15, 1. * High steam pressure in locomotive service (1907), 10. Published by the Carnegie Institute of Washington. 15034———4. 350 COX. absorbed when a low steam pressure is used, there is a limit below which one can not go, for new losses appear which more than compensate the gain. Radiation —A portion of the heat value is lost by radiation through the fire doors and furnace walls. By the use of a larger furnace and boiler the exothermic loss would be less. More favorable figures than mine have been attaimed by the Manila Electric Light and Railroad Company for Australian coal of the same source and similar composi- tion as that of tests Nos. 1 and 2, Table IL; however, it must be remem- bered that they operate their steam boilers in large units and that my figures are thoroughly representative of plants of 75-horsepower rating. Other factors ——There are many other factors which enter into con- sideration such as the physicial condition of the coal,*’ small experimen- tal errors in its use, personal variables, air leaks which dilute and cool the gases before absorption takes place, relative load carried, moisture from the air and the water of combustion which must be expelled through the stack as superheated steam, etc. Perhaps the greatest of these variables are the fireman and the moisture of the air. As a rule, the fireman is a cheap laborer secured more for his muscle than his brains, is indifferent to his work and does it in the way that requires the least energy and initiative on his part. A fireman must be intelligent or have constant intelligent supervision to obtain good results. In hand firing, instead of carefully spreading the coal or coking it and then working it back gradually, a stoker will often spread over the fire a tremendous amount of green coal. In this way the flames are smothered, the instantaneous evolution of combustible gases is out of all proportion to the supply of air, they are cooled perhaps below their ignition temperature and thus a large quantity leaves the system unburned. A deep fuel bed is called for in a producer-gas plant, but in steam boiler practice where a complete combustion is desired so that all of the carbon of the fuel will be converted into carbon dioxide a thin fuel bed is needed. When it is noticed that the steam pressure does not respond to the new supply of coal, the fireman with a slice bar or hoe will stir up the new fuel together with that already on the grate, the result being still further loss of coal. Greatly increased evaporation and saving of coal will be obtained by prohibiting these practices. The tendency of most stokers is toward a too frequent use of the bar. If Philippine coal is properly stoked it is not necessary to poke the fire at all. I have made a test of seven hours on this coal without once putting a bar in the fire box. The great difference in the moisture going into a furnace day by day, largely due to the variation of the daily humidity as well as that between the dry months and the rainy season, had often been noted; but it was left for Mr. Gayley*S to obtain definite data and show the considerable “There is a marked tendency of the coal from certain parts of the Philip- pines to fall to pieces. Care must be exercised to prevent the production of a large amount of slack in handling for it reduces the value for steaming purposes. Tron and Steel Inst. (1904), October. PHILIPPINE COALS AS FUEL. Syl economy in the working of blast furnaces by reducing the moisture in the air blast to a low and practically constant amount. It is stated as demonstrating this economy that prior to drying the air, throughout a period of eleven days the daily production of iron in the blast furnace was 358 tons with an average consumption of 2,147 pounds of coke per ton of iron, while for a period of sixteen days when the dry-air blast was used the daily production of iron was 447 tons with an average consumption of 1,726 pounds of coke per ton of iron. This shows a credit balance of 20 per cent greater output of iron and 20 per cent reduction in fuel consumed per unit of pig iron and output. However, there are other considerations. Unquestionably the greater output was largely caused by the more perfect maintenance of the regularity of the furnace owing to the practically constant amount of water in the blast. The gases in the former case were composed of 22.3 per cent of carbon monoxide and 13 per cent of carbon dioxide escaping at a temperature of 538° and in the latter of 19.9 per cent of carbon monoxide and 16 per cent of carbon dioxide escaping at a temperature of 376°, so that the economy of fuel is partly traceable to more perfect combustion and less loss through the escape of the gases. However, the fact remains that the saving through the use of dry air and the loss due to the specific heat of the moisture in the use of ordinary air is a great one, and this applies alike to all combustion furnaces. The moisture of the air is a large factor in the tropics, where the atmosphere is of almost unvarying temperature, the thermometer normally standing at 30°, and the humidity is high, the air often being almost completely saturated. The average weight of the water enterimg the furnace in the above tests was about 5 per cent of the water evaporated in the boiler. ven when all of these factors are taken into consideration there are sometimes abnormalities in the evaporative efficiency of a boiler which it is hard to explain. Some boilers owing to individual superiority, due to rapidity of water circulation, the use of water that does not foam, ete., are more efficient than others; some furnaces burn all of the volatile matter of a coal while others waste it and even the same furnace behaves differently with different coals. Theoretically, the volatile matter should be expelled from a coal on the grate and the fixed carbon simultaneously burned, thereby keeping the fuel bed intensely hot. The combustion of the volatile combustible matter should be completed in the combustion chamber. Coals high in fixed carbon burn with a short, hot, smokeless flame and combustion is nearly completed a short distance above the fuel bed, but with highly volatile coals the combustion is incomplete even at the rear of the combustion chamber. 352 COX. I have already shown*® that when Philippine coal is rapidly heated in the ordinary laboratory analysis according to the directions recom- mended by the committee appointed by the American Chemical Society,°° there is a very large mechanical loss amply indicated by the shower of incandescent carbon particles which are driven off during the first one or two minutes heating. Without the most careful stoking in the furnace there is probably the same rapid expulsion of the volatile matter as in the laboratory method, with a corresponding quantity of fine particles carried mechanically in the gas stream and to a greater or less extent deposited or burned out of the range of the absorption tubes. I have also shown*! that the presence of water serves to dampen down and hold together the solid particles of a coal, thereby preventing mechanical loss. This is probably where the advantage, if any, comes when an engineer wets a highly volatile coal. It has been shown** that fuels classified according to the increasing percentage of volatile combustible in their total combustible matter, when burned under a Heine boiler decrease somewhat in efficiency. While this conclusion holds when the number of samples averaged is sufficiently large, one must avoid too wide an application of the general- ization. Often there are physical features and special reasons for choos- ing one coal before another when theoretically it is not so good. In coking and non-coking coals apd in those entirely different physically, for example, slack and briquettes, clinkering and non-clinkering, there are factors which have many times more weight and such a generalization hardly could be applied to these, while such a comparison is perfectly legitimate and helpful to coals of the same class and physical condition. It is hoped that as soon as the public realize the availability of reliable information regarding coal, both concerning its composition and steam- ing value, these means of determining its value may be more often resorted to and that guesswork may be eliminated from the purchase of a coal. SUMMARY. The object of this investigation was to determine the steam-making value of the coals of the Philippine Islands as compared with the foreign coals offered on the market in this Archipelago. All the tests which are described in full were made at the Bureau of Science with a 75-horsepower water-tube Babcock & Wilcox steel boiler over a hand-fired furnace. An average of 1114 per cent of the rated capacity and an ayerage steam pressure of 7.4 kilograms per square Cox, A. J.: This Journal, Sec. A (1907), 2, 43. © J. Am. Chem. Soc. (1899), 21, 1116. §'Cox, A. Je: Loe ncit. 59! =U. S. G. S. Bull. (1907), 325, 89. PHILIPPINE COALS AS FUEL. B50) centimeter (105 pounds per square inch) was maintained. 'The average length of the tests was about seven hours. ‘The plant, the apparatus used and all conditions were preserved as nearly constant as possible. It was my purpose to burn each coal with the maximum economy in this type of furnace. For a Philippine coal a regular and uniform method of firing is essential. It was found that the best method of firmg was in small quantities every four or five minutes. A thin fuel bed is also needed and it must not be frequently worked. An entire test of seven hours duration was made without once disturbing the fire. Inert matter in a coal is detrimental to its value in that the total number of heat units is proportionally decreased. Moisture further reduces the efficiency directly by the specific heat of the water, but the content of ash ordinarily found in Philippine coal has very little if any further effect. It seldom produces clinker and for this reason the presence of sulphur is no detriment. Moreover the percentage of sulphur in Philippine coal is usually extremely small. A short fire box, the usual vertical baffling and an ordinary bar grate are not suited successfully to burn Philippine coal. An average of 94 per cent less of the theoretical heat units were absorbed by the boiler when Philippine coal was consumed in the plant of this Bureau than with the Australian coal ordinarily used and for which the plant was selected and installed. The efficiencies recorded in Table II include those of the boiler, fire box and grate. There is very little variation in the steam pressure and the amount of water evaporated per hour. When a boiler with a satisfactory rate of water circulation, absorbing surface, etc., has been used the deviation from the maximum efficiency of a plant depends largely on the adapt- ability of the furnace grate and stack. The economy is greatest with those coals which have a high fuel ratio, burn completely and give a high combustion chamber temperature. With satisfactory absorption the greater the difference between the temperature of the combustion chamber, gases and the boiler, the greater the efficiency and the less the loss to the stack. When Philippine coals are burned in an ordinary furnace they are at a disadvantage as they tend to burn out of the range of the boiler tubes with the result that there is low evaporation and high chimney temperature. A longer fire box or an increased number of baftle walls, or both, and a carefully selected grate would probably greatly increase the efficiency of Philippine coals. If the number of baffle walls is greatly increased, care must be exercised that there is sufficient draft. The tendency to burn out of the range of the boiler tubes which coals high in volatile matter show, is aggravated by an excessive draft. The greater the quantity of air drawn through the fuel bed, the more rapid the combustion and the farther in the rear of the combustion chamber it takes place. With a heavy draft the result is high chimney temperature 354 cox. and low efficiency. On the other hand, too little air results in low efficiency due to incomplete combustion. Highly bituminous coals deposit much soot which may reduce the efficiency of the heating surface, and the formation of scale is a factor which needs close attention if maximum efficiency is to be attained. With a change in efficiency other factors of the heat distribution also vary. The radiation is especially variable with the size of the plant and the temperature of the combustion chamber. The size of the fuel is a-very important factor. The crumbling of coal reduces its value for steaming purposes. There is a tendency of coal from some parts of the Philippines to fall to pieces. Care must be exercised in handling to prevent this. _ The moisture of the air is a large factor in the tropics. With an evenly warm, almost saturated, atmosphere the amount of water entering the furnace is enormous and considerably lowers the capacity and ef- ficiency of the plant. ; The average of the calorific values of all the Philippine coals tested is 6,003 ** calories and that of the Australian coal ** purchased by the Government and furnished to this Bureau for fuel is 6,614. In in- dividual cases the calorific value of Philippine coal is as much as that of the Australian coal and in one case showed an efficiency in this plant, which is unfavorable to Philippine coal, within 3.75 per cent as great as that attained when the Australian coal was fired. With respect to ash, clinker formation and the production of smoke the Philippine coals are superior to any others offered on the Manila market. 9/5 calories=B. T. U. ®t This coal was tested in June, 1907 (tests Nos. 1 and 2, Table IT). ILLUSTRATIONS. Priate I. Babcock & Wilcox boilers used in making the tests (cf. p. 304). Il. Voltmeter and ammeter diagrams of tests numbered 1, 2 and 3, Table IT (p. 311). : III. Voltmeter and ammeter diagrams of tests numbered 4, 5 and 6, Table II (p. 311). TV. Voltmeter and ammeter diagrams of tests numbered 7, 8 and 9, Table IL (p. 311). V. Voltmeter and ammeter diagrams of tests numbered 10, 11 and 12, Table If (p. 311). VI. Voltmeter and ammeter diagrams of tests numbered 13, 14 and 15, Table II (p. 311). VII. Voltmeter and ammeter diagrams of tests numbered 16, 17 and 18, Table IL (p. 311). VIII. Charts used in judging the color of the smoke (cf. p. 310). IX. Grating of the charts in Plate VIII drawn to the exact scale. X. Figure showing graphically the steam-pressure gauge readings of tests numbered 1 to 10, recorded in Table IV, A to J, inclusive. The dotted curves are supplemented from automatic indicator diagrams in order to show the maximum and minimum yariations. XI. Figure showing graphically the steam-pressure gauge readings of tests numbered 11 to 18, recorded in Table IV, K to R, inclusive. The dotted curves are supplemented from automatic indicator diagrams in order to show the maximum and minimum variations. XII. Figure showing graphically the temperature of the flue gases, base of stack, of tests numbered 1 to 10, recorded in Table IV, A to J, in- clusive. XIII. Figure showing graphically the temperature of the flue gases, base of stack, of tests numbered 11 to 18, recorded in Table IV, IK to R, in- elusive. Page. Fic, 1. (In text.) Showing the flue-gas sampler used in drawing the gases Loven ally SUS eee eee see oe eee. eee ee eee Seen ieee a 308 2. (In text.) An ideal section showing an ordinary type of boiler with an elongated fire box and additional baffle wall................-..---....... 344 a os 5. III, No. [PHIL. JOURN. Sci., VOL. Cox: PHILIPPINE COALS AS FUEL.] PLATE Cox: PHILIPPINE COALS AS FUEL.] [PHIL. JourRN. Scr., Vou. III, No. 5. { | Chart Cox : PHILIPPINE COALS AS FUEL.] (PHIL. JourN. Sct., Vou. III, No. 5. PLATE Ill. Cox: PHILIPPINE COALS AS FUEL.] [PuHiIL. JourN. Scr., Vou. III, No. COX: PHILIPPINE COALS AS FUEL.] [PHIL. JouRN. Sci., Vou. III, No ie S. So = wa Ta orn PLATE v. S PHILIPPINE COALS AS FUEL.] [PHIL. Journ. Sct., Vou. III, No. 5. PLATE VI. Cox: PHILIPPINE COALS AS FUEL.] (PHIL. JourN. Sci., Vou. III, No. 5. PLATE VII. "MIA Stw1d Qo Sct 20u Vow ‘G ‘ON ‘III “IOA “10g ‘N¥nOf¢ “11H q] : [Ia0q SV SIvOO ANIddITIHg : X09 aBTE fe ie vi ela Dias ve | cf “ Mendez Nunez x é Q pale pene ener eet oat Sy G Nagcarlang\y — Lilio Sto.Toma: Ly Tanalan_ -Alaminos planauan. <] Kilometers ‘ 10 5 ° 10 20 30 % Clay deposits Map oF LAGUNA DE Bay SHOWING THE REGIONS FROM WHICH THE CLAY WAS TAKEN. LAGUNA CLAYS. By Axyin J. Cox. : (From the Laboratory of Inorganic and Physical Chemistry, Bureau of Science, Manila, P. I.) This investigation was undertaken at the request of the Bureau of Education to enable it to choose a clay for use in the pottery school, the building for which is now being erected in Santa Cruz. None of the clay deposits of Laguna Province have been thoroughly studied up to the present time and, therefore, I will first discuss those so situated that the stripping of the material would entail the least labor and the soil overlying could easily be disposed of. Such clays would be economical to use were they of high grade. A paper on the clays from the Island of Luzon has already been published in which the uses and the chemical and physical behavior of some of the Laguna clays and the influence of the fluxes were thoroughly discussed. 'The data there given are to a great extent directly applicable to the samples treated in this paper and the application is so simple that the interested reader may make it for himself. The following statement regarding kaolin may assist in the proper interpretation of the results given below. The composition of pure kaolin (kaolinite) calculated from the theoretical formula Al,0,.2Si0,.2H,0 is— Per cent. Silica (Si0,) 46.65 Alumina (A1,0;) 39.45 Water (H,0) ; 13.90 Total 100.00 Examinations of kaolins from Harris Company, Webster, North Car- olina,? and from Glen Allen, Missouri,’ are as follows: Cox, Alvin J., This Journal, Sec. A., (1907), 2, 413. 2N. C. Geol. Sur. (1897), Bull. 13, 59 et seq. *Mo. Geol. Sur. (1896), 11, 578 et seq. 378 COX. Chemical analyses. [Figures give percentages. ] ; Fluxes. Water (H20). a eS 4 = Ske Ch | a Oo. ve) he a |r EIS : . |ael ¢ Source. 6 Sie Lo See | Sy es ei ee x = 3 O79] 0 a CS) a toad ri = PAE] S 2) 2 \os|s$2|2 lee|2]a4)s |e l4a=| = eG) geal HE SS ees eis St a om 3 u- | a ‘=| a oc ° = S = = 5) oO =) ey fo) o 2 o 4 is} Dn < & & A a n (Wj <4 Q i a Crude kaolin Webster, N. C -_| 62.40 | 26.51 | 1.14 |---_-- 0.57 | 0.01 0. 98 8.80. | 0.25 |------ 2.70 Washed kaolin, Webster, N. C __| 45.78 | 36.46 | 0.28 | 1.08 | 0.50 | 0.04 0. 25 13.40 | 2.05 |------ 2.15 Glen Allen, Mo --_| 72.30 | 18.94 | 0.40 |__--_- 0.68 | 0.39 0. 42 a ea ae 1,91 Physical properties. Tensile strength Water of the air-dried | Shrinkage (per cent). added to sample. ane! Source. Wire ane paste a (per cent Kiloper Pounds __ | Fire at total Seats @ Oates Air. cone | Total. weight). meter. inch. Cc enh oC Webs teriNi Ciaoses Se eee eee 42.0 1.55 | 22 6 4 10 GlenwAlllen; so sess eerene nena meen 23.2 0.84 | 12 4 8.4] 12.4 The chemical composition of the clays given below shows them as they occur and may include impurities that may be removed by washing, sieving, etc. : e MATIQUIO. The outcrops appear at various places along the Butadero River northwest of Matiquio. Three outcrops, which have been staked by Laguna Province, are about an hour’s walk from Matiquio. They are very close together, the prismatic compass shows them to be 13° west of north of Pila and the aneroid barometer shows their altitude to be 240, 300 and 270 meters, respectively. The natives report that formerly 1,000 five-kilo baskets were taken from this point and shipped for use at Bilibid. All three of these outcrops are on the steep hillside, perhaps 40 meters above the river. The clay slopes wash down easily and one outcrop was so obliterated that it was impossible to obtain a sample. The other two were readily uncovered and show that most of the thoroughly weathered clay has been removed, or that the material has not yet been completely weathered. Most of the post-auger holes gave very gritty and imperfectly weathered samples. They all contain pyrites, which will prevent their burning to a white product. LAGUNA CLAYS. 379 A light gray sample of fair quality was taken from the lowest outcrop. Three determinations gave its tensile strength as 4.8 kilograms per square centimeter (68.3 pounds per square inch) and its air shrinkage as 1.4 per cent. Continuing up the Butadero River, on the left there is a large outcrop of clay about 3 meters wide and 4 meters high just above the high-water line. On the surface it is bleached to a whitish color, but a few centi- meters below it is blue. This clay contains pyrites and grit which unfit it for pottery use in the raw state. A short distance farther up, the Sunka Sunka and the Kaloong Rivers join to form the Butadero. About a half kilometer up the Kaloong on the right is a perfect illustration of the breaking down of the basal rock in the formation of residuary clay deposits, but the decomposition of the deposit has not continued far enough to form any considerable quantity of good clay. On the ridges the decomposition is very incomplete and the clay is full of unweathered rock fragments and pyrites. In the water channels and moist places the decomposition is fairly complete, but these areas are neither large nor numerous. However, a sample from one of these places was the best that I found in the Matiquio region, and when subjected to laboratory examination gave the following results: Chemical analysis. [Figures give percentages. ] Fluxes. Total Water | Tita- Silica Ala iron pees (H.O) | nium} Sul- | Total (SiOz). (Al,0s) given | Time Mag- | goda | Potash tees below | oxide | phur. | fluxes. Oe a8 | (Gad). | DES!® |(NasO).| (KoO) “| 110°. | (TiOs). ferric | (MgO). AG NO oxide (Fe203) 61.00 | 19.71 3.48 0. 34 0.07 0.23 0. 61 8.18 5.97 (0895) | aan 4.73 b64,.8 | 20.96 | >3.70] »0.36 | 0.07 | »0.24 | >0.65] »8.70) »0.00 | >1.01 -| 5. 02 Mostly water above 110°. b Recalculated free from water below 110°. Physical properties. Tensile strength. Shrinkage (per cent). Color. Water | Air dried, 3de-| Burned, 4 de- added to | terminations | terminations give a averaged. averaged. workable k paste Fire, ; ance Burned with (per cent | 4 z Air. jat cone) Total. | Air dried. free access poet ee Pounds' a Pounds) No. 9. of air. weight). per per ee square eer square meter. | CH: | meter. | ich- 22.4 3.47 49.3 9.07 129 4.7} +0.4 4.3 | Light gray ___| Light gray. 380 cox. The iron content of this sample is so large that the clay could not be expected to yield a white product, but its physical properties, particularly its plasticity, tensile strength and shrinkage are such as to make it useful in the manufacture of colored wares. The clay does not crack in burn- ing. In the future these deposits will probably be valuable, but the outcrop indicates that the supply at present would be quite limited without a laborious and expensive system of troughs and tanks. Such operations would not pay except on a clay of the highest quality. A limited amount of this clay could probably be mined and especially because of its low shrinkage would be used to recompose other clays which in some ways have more desirable physical properties. MAJAYJAY. The deposit at Majayjay is on Mangulila Creek about 30 or 40 meters from its mouth where it empties into the Dalatiuan River at an altitude of 180 meters. It is 300 or 400 meters above the point at which the Majayjay-Lucban trail crosses the Dalatiuan gorge. There is a large amount of clay in sight. On the left side of Mangulila Creek the clay is exposed in a bank 15 or 20 meters long and about 4 meters high. This material is probably of sedimentary origin. I removed the surface and thoroughly investigated the clay. It is very uniform in quality and plasticity, free from grit and bluish in color when wet. The exposed surface is yellowish, unquestionably because of the fixation as oxide of the large amount of iron which the clay contains. The deposit appears to extend under the stream and outcrops on the right bank in a much whiter condition, but under large boulders. Samples of the deposit were carefully taken on both sides of the river and investigated in the labo- ratory. The two were found to be almost identical in their general characteristics. The data and results are as follows: Chemical analysis of the clay from the left bank. [Figures give percentages. ] Fluxes. pk Alu. | Total Loss_| Water | Tita- Manga- Silica vaitreyay |} seein onigni-}| (H20) | nium | Sul- nese (SiOz). (AloOg)| given i Mag- Soda | Potash tion. below oxide | phur. | oxide | Sod eas Lime | nesia (Na.0).| (K 0 ; 110°. | (TiOs). (MnO). ferric | (CaO). | (MgO).|(N820).| (K20). | oxide (Fe.0x) (By difference.) | 39.55 | 28.41 | 10.27 0.35 0.33 0.17 13. 62 6.25 1053 Seen trace. b42.15 |>30.3 | 10.94] 0.37 | >0.35 >0.18 b14.50 | >0.00 | 1.12 |___-_-_- > trace. ® Mostly water above 110°. > Recalculated free from water below 110°. LAGUNA OLAYS. 381 Physical properties of the clay from the left bank. Tensile strength. MEO Shrinkage (per cent). Color. added to - . givea Air dried. Burned. workable paste " c (per cent | Kilos [pounds E08 |pounas total per per per per i : 2 4 weight). | Square) .iyare | Sduare| are Air. Fire. Total. | Air dried. Burned. centi. centi- | % meter, inch meter ane: 37.0 1.9 227 | none. |’none. 8.65 17.5 | 26.15 | Yellowish-gray___| Red. "8 8 of the 9 briquettes were so badly cracked that they were not usable. > Cracks open in all directions and the briquettes warp badly. Physical properties of the clay from the right bank. Tensile strength. ae? Air dried,2de-|_, Shrinkage (per cent). Color. = terminations Burned. givea a workable | ®#Vveraged. | ( paste | per cent ae : \ total Kilos Kilos eel weight). | Per pounce per negeee F ; easy Burned with | square square Air. | Fire. | Total. Air dried. free access centi- | S4U8Te | centi- | Sauare of air, inch. inch. meter. meter. 30.0 1.55 22 | none. |2none. 8.5 14.8 PERSBS MCHC ao a Red. 8 Cracks open in all directions and the briquettes warp badly. The red color of the burned product and the abnormally great fire shrinkage indicate that this clay is worthless for the manufacture of fired products, furthermore the product cracks badly upon air drying. BOTOCAN. There is a peculiar ridge of red clay which is said to extend nearly all the way from Lucban to Pagsanjan. As it exists in such large quantities I thought it might be of commercial value and a small sample was taken about a hundred meters south of the barrio of Baquio. Its analysis and physical properties are as follows: Chemical analysis. [Figures give percentages. ] Fluxes. ” Water Tita- | Manga- flies Ea arora b, ee (H.0) | nium |” nese 1Q9). 1ron glv- : agne- . elOw oxide, oxide (A1,03). enlaeiten! Lime sin Soda Potash tion.® 110° TiO MnO ric oxide| (C®0)-| (argo),| (Na20). | (K20). Po NS EE) (Fes0,). 34.55 | 28.73 | 12.77| 0.00] 0.16] ‘small. | small.| 13.02] 10.09} 117] 0.06 38.4 | 31.9 | »14.2 | »0.00| >0.18| »small. | »small.| »14.47| »0.00| »1.30| »0.07 * Mostly water above 110°. 75034—_6 > Recaleulated free from water below 110°. 382 COX. Physical properties. ] | Tensile strength. Shrinkage (per cent). Color. Water added to . - give a Air dried. Burned, workable paste Kilos Kilos Burned with free er cent i i i i \ (Denes per Hounds ner Houtas Air. | Fire. | Total. | Air dried. |“ 5 acess of air. eight). | Square | square welehy) centi- leptons centi- eee = meter. meter. 2 8654) 0815) |) > 2227/2282 | Bees 14.0 17.5 31.5 | Dark red _| Very dark red. This clay on burning gives a very dense, hard body and a good paving brick might possibly be manufactured from it by recomposing it to increase its tensile strength and overcome its cracking in the air. MOUNT MAQUILING REGION. The clays of this region are the only true kaolins that I have seen in Laguna Province. The knowledge of their existence dates back to Spanish times, but their extent is not known. The outcrops in all of these deposits have been very much worked over and in such a manner that it is not easy to obtain very many particulars concerning them. The laborer digs a hole through the overburden just large enough to admit his body ; when he reaches the kaolin he burrows it out in every direction _ as far as he can without danger to himself. When he has exhausted one hole he goes a little farther and digs another. Sometimes the strippings from the second hole have been thrown into the first and even when this has not been done, since the deposits are in all cases on the mountain side, slides and washes have filled all the old workings and it is impossible to estimate either the amount of overburden or the extent of the deposits. Only systematic boring or the uncovering of a large area can determine this. LOS BANOS. This deposit is 6 or 8 kilometers from the lake, southeast of Los Banos on the mountain side above Bagong Bola Creek. This is the only deposit known to the natives and is reached by a very poor trail. The highest point on the trail is about 370 meters, while the deposits are about 350 meters above sea level. The natives say that no clay has been taken from this point since Spanish times. One man informed me that he formerly made excavations here and that many of the pits were dug to a depth of five meters. There are scores of holes partially or almost completely filled by wash ; their existence indicates that the deposit extends over a considerable area. In several places there are appearances of an outcrop, but closer examination shows them to be only excavated mate- rial cast aside by the miners and proves that no superficial examination can possibly reveal much regarding the thickness or extent of the deposit. I opened up two of the old pits which I thought perhaps representa- 383 LAGUNA CLAYS. tive. I excavated one to a depth of over two meters and then bored on a slant in two directions with a post-hole auger for 15 meters more. On one side the material was gritty and only partially weathered, while on the other a very good, uniform, sample of kaolin was obtained. The latter was investigated in the laboratory with the following results: Chemical analysis. [Figures give percentages. ] Fluxes. | Water| Tita- Manga- Silica elu Jota Benen (H20)| nium | nese | Total (Si02)-| (41,03)|givenas| Lime |M®8M°) soda | Potash| tion.» |PEOw| oxide | oxide | fluxes ~ ferric | (CaO). | eS! |(NaoO).| (Ko0). TLOS S| MLLO2)) (MnO): oxide (MgO). i (Fe203) 49.42 | 30.45] 1.61] 0.00] 0.21] 0.07] 0.09] 13.72] 5.86] 1.11] trace.| 1.98 »52.45 | 32.3 | >17.1| >0.00) >0.22] 0,07 0.10 | »12.48 0,00 | 1.18 |>trace.| >2.10 ® Mostly water above 110°. > Recalculated free from water below 110°. Physical properties. a Water Tensile strength. Shrinkage (per cent). Color. added to give | awork-| Air dried. Burned. able paste 5 A Fire at i (per | Kilos fpounds| Kil0S |pounds| air. | cone | Total.| Air dried. | Burned with free cent per per per per No. 9. access of air. total | Square) cquare | $date | square weight)| centi- inch centi- inch meter. meter, 4 38.6 1.56 22.2 4, 22 60 5,2 6.3 11.5 | Creamy white| Light creamy 3 white. ® Does not crackle. The other pit excavated to a depth of a meter and then bored on a slant in one direction yielded quite a different sample. This contained a small amount of pyrites and when moist was of a bluish tint. The natives say that material of this quality formerly overlay the whole deposit of white clay. The laboratory data and results on this sample are as follows: - Chemical analysis. [Figures give percentages. ] Fluxes. Water Tita- Silica | Alumina pote Ree (H2.0) nium (S102)-: | (41205). given as ame Mag Hoda potest Hontay aoe ows (Hoss erric aO). a0). x0) 4 2 oxide (MgO). ‘ ) ES) (Fe203). 43. 83 31. 86 5. 86 0.14. 0.11 small small. 15. 04 2.71 0.80 b 45.0 032.7 66,03 | 0.15. bO.11 bsmall | >small. v15. 44 b0.00 b0. 82 8 Mostly water above 110°. > Recalculated free from water below 110.° 384 COX. Physical properties. | Water Tensile strength. Shrinkage (per cent). Color, added to give a work- Air dried. Burned. able paste a = ‘Burned with (per Killos Pounds Kilos Pounds| Air. | Fire. | Total. Air dried. free access cent DEE er REL per of air. total | Square) onare | SAUare | square weight).| cemti- | Sno, | centi- | Sach meter, meter SB n 0 |e neces | oars 8.44 120. 4.65 4,25 8.9 | Light grayish- | Bluish-white. white. 8 Crackles considerably. These three samples taken together may indicate that the material is still in the process of formation and that formerly only the pockets of the best clay were sought and worked out. Both the chemical and physical characteristics of these two samples indicate them to be suitable for use in the manufacture of a good class of pottery, but probably a large amount of sorting would be necessary in the mining of the clay. CALAMBA. Two kaolin deposits are known in the region of Calamba; one in the Pajo Canton and the other below Point Alipasio, overlooking the Santo Tomas Road. A pony trail on the right side of the cafion leads to the former; it is steep in places but fairly passable in the dry season. This deposit is about two or three kilometers from the lake and at an altitude of about 200 meters. There are outcrops on both sides of the cafion, but like those of Los Bafios, they have been disturbed until definite information regarding them is difficult to obtain. The natives say that on the right side of the cafion the clay is the whitest that has ever been found. I made several borings at spots indicated by men familiar with the deposit. Some samples were white, streaked with black, perhaps vegetable matter, but one was uniformly white. Since it is claimed that the latter was not an unusual pocket I have studied the sample in the laboratory, obtaining the following data and results: Chemical analysis. | [Figures give percentages. ] Fluxes. | Ewe Alu- P Water Tita- Giics | mina | Tec Loss On| (a0) | ninm | Total | *|(Al00). piven as] Lime |MUERC-| soda | potasn | tion | igs | Go, | B= erric a0). 0). 20). i atias (Mgo). | (N#20).} (K20) | | (Fe.0s) . ¥ | | aa | 55. 99 28.77 0.89 0.18 0.03 0. 08 0.09 11.59 2.42 0.91 1.27 >57.4 | 29.47 v0. 91 >0.19 %0,03 >0. 08 »0.09 | >11.87 »0.00 >0.93 | 1.30 ® Mostly water above 110°. > Recalculated free from water below 110.° LAGUNA CLAYS. 385 Physical properties. | Water Tensile strength. Shrinkage (per cent). Color. added | ao ee | Air dried. Burned. able ‘ | paste A | é Fire, at : Burned with (per ae [Pounds ae Pounds| Air. cone | Total. Air dried. free access cent equiv MIECUE square |eernl No. 9. of air. total re | square re | square | . weight).| Cent: | “ineh centi- Sraolh “ | meter. meter | - | | | 36.3 271} 38.6 | 3.5 | 50 |} 3.06 7. 64 10.7 | Light cream. Light creamy- | | white. *® Burns excellently. The deposits were being worked on the other side of the Pajo River at the time of my visit; the product is sacked and carried on the backs of natives or ponies to the barrio of Bukal on the lake shere where it is made into balls and shipped to Manila. The material as mined is more or less streaked with red, but when the less plastic lumps are discarded and when macerated it works up to a light cream and is bought by the Chinese of Binondo, who make of it a sort of whitewash which is said to be better than the whiter product above mentioned, perhaps because of its greater tensile strength. Certain of the physical proper- ties, especially the behavior under fire, have no significance when a material is used for covering walls. The open workings indicate a considerable amount of this class of clay. Two of the pits were sampled. The lowest opening on the mountain side was carefully cleaned out to a depth of at least two meters and then I took an average sample of the material, which I removed from a meter bore-hole in the bottom. The clay from this pit was accepted by the natives without sorting. The data and results of the laboratory tests of the sample are as follows : Chemical analysis. [Figures give percentages. ] Fluxes. a Alu- Water | Tita- Silica | mnina Hotel ee (H.0) | nium | Total (Si02). | (A1,03).| given as| Lime |M@8e-| soda | Potash | tion.» | Clow | Oxide | fluxes. y i =I 4 2, TiOs). ferric CaO). presley NasO).| (K20). HY ( s oxide ‘ (MgO). ‘ dal ) (Fes0s). 42, 23 37. 32 1.41 0. 23 0.07 0.11 0. 35 15, 84 1.92 1.00 2.17 >43.1 | »38.05 b1.44 ’0. 23 >0. 07 >O.11 | »0.36 | >16.14 60.00 b1,02 | »2.21 L | 8 Mostly water above 110°. » Recalculated free from water below 110.° 386 COX. Physical properties. Tensile strength. Shrinkage (per cent). | Color. | | Waters enna j | = r | added | 4 jr dried, 4 de-| Burned,*3 de- | Wee, terminations | terminations | lane averaged. averaged. | paste — Fire at = ae Burned with | per =F =o Air. cone otal. Air dried. Tee access | cent | Kiles | pounas| KOs | pounds No. 9. of air. woe Ss aes ne eure | rS quare pveieht).| conti. sana | centi- ae meter. * | meter. ¥ | 25.4 3.78 53.8 4.00 57 3. 06 11.1 14.16 | Creamy-white. | Rose-cream. | * Crackles to a certain extent. Another and larger opening about ten meters from this and about three or four meters higher was also sampled. This pit had been bur- rowed out in every direction to a depth of 2 meters. I excavated a third meter and bored still a fourth to obtain my sample. It was an average of the run of the deposit and showed more red streaks than the former sample and had an extremely mottled appearance. In this pit the natives sorted and discarded the less plastic lumps. The tensile strength of my sample will be seen to be lower than that obtained from the other pit, but it is believed that the clay removed by the carriers was of about the same quality. The data and results of the laboratory tests are as follows: Chemical analysis. [Figures give percentages. ] | Fluxes. | | | | | | Water Tita- | Silica | Alumina) Total onieni- | (H20) | nium | | (SiGe) | (Al20s). | given Gas Mag- eng Howsh tion.® below Giey | | as ferric! (CaQ). Nao). | (Ks). 2h | oxide (Mg0).| | | (Fe20s). | | | | 43. 28 37. 85 3.39 0.08 0.04) small. | small. 14.2} 0.89 25 43.7 >38.2 3.4 60.08 | »0.04) small. | small. | 14.3 | 60.00 | 1.25 | 1 | * Mostly water above 110°. > Recalculated free from water below 110°. LAGUNA CLAYS. 387 Physical properties. Tensile strength. Shrinkage (per cent). Color. Water a = added | Air dried, 3 de- | to give | terminations Burned.* a work- averaged. i | able , 5 | paste nA ire ateten Nieahea Burned with tl 1%; =A ir. |atcone) Total. ir dried. ree access (pecan aes Pounds ies Pounds No. 9. of air. weight). ) square|_ Pe’ | square} . Pet ; centi- Nae centi- eee meter. > | meter. 7 26.6 1.64 | 23.3 2.95 42 1.6 9.6 11.2 | Cream -_______ Pale lilae. 4 Crackles to a certain extent. Some of the natives state that they have made excavations as deep as four meters, but although they worked regularly in these deposits they showed such timidity about entering the pits that I doubt if they often go deeper than the holes which were open during my visit. The information gained from the chemical and physical tests of this deposit indicate the usefulness of this material for making certain grades of pottery. The deposit below Point Alipasio in the region of Calamba is at an altitude of 210 meters and has apparently been very little, if ever, worked. It is on the dry side of the mountain and the weathering has taken place very much more slowly; moreover the deposit is quite hot and the heat soon dries out the water which is perhaps the chief desintegrating factor in a tropical country, this prevents further action. In some places the basal rock is scarcely decomposed at all, at others there is a thin layer of completely weathered, grayish clay, while in still others there are pockets probably of no great extent, of a good quality of kaolin. One of these at the base of the deposit was sampled and investigated in the laboratory. The data and results are as follows: Chemical analysis. [Figures give percentages. ] Fluxes. Water | Tita- | Sites || Ae | Boe poss.on (H.0) | nium | ‘Total | (Si02). | (41,0,).\given as| Lime Mae. Soda | Potash | tion.» pow Gas fluxes. oxide | (CaO). | REA |(Na.0).| (K20). : 2). | ferric (MgO). | (Fe,03). | = | Ps \ 43.16 38. 64 1.19 0.09 0.14 0.08 0. 02 14, 55 1.42 1.54 | 2 b43.77 | »39.20 »1.20 »0.09 b0, 14 »0.08 60,02 | »%14.75 »0, 00 b1,56 | %3.53 a Mostly water above 110°. » Recalculated free from water below 110°. 388 COX. Physical properties. Tensile strength. Shrinkage (per cent). Color. Water | added | Air dried, 6 de- to give | terminations Burned. a work- averaged. bl meats F Or Meena Buraed with Wes 3 Air ire. otal. ir dried. ree access (reniceny ellos Pounds Dae Pounds of air. weight). : Bbt)- | square | cauare | SUAte| square meter. inch. meter. 34.9 1.81 25.8 1.9 27.1 6.1 13.3 19.4 } Bluish-white _| Cream. | | The physical properties of the grayish clay are as follows: “ Crackles badly. Tensile strength. Shrinkage (per cent). Color. Water |— ; aos Air dried. Burned.* | a work- | able paste i rai Fire Burned with (peRcent eee Pounds ue Pounds} Air. |atcone) Total.| Air dried. free access a : weight). | S9uare) . are SAUaTe Ks Trane ae een | | centi- ae centi- aa meter meter. ‘ — — i = 33.4 1.75 24.9 2.39 34 4.63 14,2 | 18.85 | Grayish-white_| Cream. i 4 Crackles badly. Still another very light-gray sample was taken at a distance of 30 or Three determinations gave its tensile strength as 2.4 kilo- grams per square centimeter (34.2 pounds per square inch) and its air 40 meters. shrinkage as 2.4 per cent. At some future time the disintegration of this deposit will have produced a sufficient quantity of clay to justify its use for the manu- facture of po ttery. There is a Japanese in Los Banos who is successfully molding various He has very primitive ap- pliances and his products are all underburned, but his work indicates that with proper handling, satisfactory results probably could be obtained with this clay. The indications are that all of the high grade clays of Laguna Proy- kinds of small objects from Calamba clay. ince are more or less mixed with clay of poorer quality. It is a question what percentage of the mixture is useful and whether the expense of sorting will not be so great that it can not compete with kaolin from other sources. ILLUSTRATION. Facing page— Map of Laguna de Bay showing the regions from which the clay was taken. 377 389 VOLCANIC TUFF AS A CONSTRUCTION AND A CEMENT MATERIAL. By Atvin J. Cox. (From the Labortory of Inorganic and Physical Chemistry, Bureau of Science, Manila, P. I.) In the Philippines the varieties of stone which are known to be good for building and construction purposes are not very numerous. Moreover it is unfortunate that those that exist are not conveniently located for use. Since the first cost is the controlling factor in the purchase of materials for construction, a very poor substitute must often be used. The metamorphic and igneous rocks, capable of resisting heavy strains and weathering, which have been noted on many of the Islands of the Philippines by the early explorers,’ and by Becker,? McCaskey and Ickis,* Eveland,* Smith,’? Ferguson * and others of recent time, embrace: A. Metamorphic : Breccias, gneisses, schists and serpentines. *Meyen and Itier, J.: Reise um die Erde (1835), 2, 237; Itier, J.: Bull. Soc. geog. Paris, 3d. (1845), 5, 365 et seq.; Roth, J.: F. Jagor’s Reisen in den Philippinen, Berlin (1873), 344; Centeno, José: Memoria geologico-minera de las Islas Filipinas, Madrid (1876), 19; Von Drasche, R.: Fragmente zu einer Geologie der Inzel Luzon, Vienna (1878), 20 et seg.; Oebbeke:. Neues Jahrbuch, Beilage (1881), 1, 495-8; Renard, A. F.: In the Report of the Exploring Voyage of H. M. 8. Challenger, London (1889), 2, Pt. 4, 160-175; Abella, Enriqué: El Mayon, Madrid (1885) ; El] monte Maquilin, Madrid (1885); La Isla de Biliran, Madrid (1885) ; Rapida descripcion fisica, geol6gica y minera de la Isla de Cebu (1886), 96-101; Descripcion fisica, geolégica y minera en bosquejo de la Isla de Panay, Manila (1890), 97 et seq.; Manantiales minerales de Filipinas, Manila (1893), 16, 18, 31, 75, 144, ete.; Terremotos experimentados en la Isla de Luzon, ete., Manila (1893), 32 et seq.; ete. Dr. Becker gives a bibliography of the early books and papers on Philippine Geology, U. S. G. S. 21st Ann. Rep. (1902), 594. ? Becker, G. F.: loc. cit., 561. * McCaskey, H. D. and Ickis, H. M.: Sixth Ann. Report of the Mining Bureau, Manila (1905), 57. *Eveland, A. J.: Bull. Min. Bur., Manila (1905), 4. *Smith, W. D.: Jbid., 5; This Journal (1906), 1, 203, 617, 1043; Ibid., Sec. A., 2, 145, 253. * Ferguson, H. G.: This Journal, Sec. A. (1907), 2, 407. 391 392 COX. B. Deep seated igneous: Diabases, diorites, gabbros, granites, quartz porphyry (dike rock), peridotites, pyroxenites, tonalite, and syenites. C. Volcanic: Andesites, basalts, rhyolites, dacites and trachytes. Some of these are merely enumerations, for no microscopic or special study of them has been made. On the other hand it is probable that the known quantities are but a small proportion of those that actually exist because of the meager exposures in the mountains and ravines, many outerops being entirely obliterated by covers of soil and heavy tropical vegetation, and because certain districts have not yet been explored. It would be out of place here to discuss the distribution of these rocks. However, it might be mentioned that the diorites and the andesites of the second and third classes, respectively, are the most generally distributed. The former occur very extensively in the Cordillera and the latter in the Mariveles district, which forms the northern headland of Manila Bay. The latter district is now being extensively mined by the Atlantic, Gulf and Pacific Company for use in connection with the sewer construction in Manila. Limestone occurs widely distributed practically throughout all the Islands. Goodman‘ reports limestone at Dumalag and Pilar, Province of Capiz, ““appear- ing especially compact and seemingly fit for a building stone. In fact the church at Dumalag, which was erected in 1873, as well as other buildings in the town are constructed of this material. I noticed, however, that some of the blocks in the church showed marked effects of weathering. Although fairly compact, I do not believe that the Dumalag limestone is any more so than that which may be quarried near Binangonan on Laguna de Bay, this latter having the further advantage of being much nearer to Manila and to tide water.” Mr. Goodman also reports with regard to tonalite that: “At the foot of Mount Anjaoan, which is just off the main road and about a quarter of a mile southwest of Colasi, is a massive outcrop of the tonalite which Abella reports as occurring in this neighborhood. At Mount Deluca, a low hill about one hundred feet high and about one and a quarter miles west of Anjaoan, the tonalite takes on a grayer color, but in texture and mineral composition remains the same. The same rock occurs again at Mount Jilonoe which is about the same height as Mount Deluca and about one and a half miles west of it. Outcrops of tonalite are encountered in various other places along the road between Colasi and the barrio of San Fernando, and there in but little doubt that practically the same formation continues for at least three miles east of Anjaoan. A thick growth of cogon covers these hills so as to expose but comparatively small surfaces of the rock in place, and these all appear massive, no systematic jointing being observable. The stone is uniform and close grained, compact and hard, takes a good polish and seems well adapted for use in heavy construction. It is well located too with a view to quarrying and transportation. The harbor at Colasi, however, did not strike me as nearly so perfect a one as I had expected and hoped to find it.” ‘An unpublished report made to the Chief of the Division of Mines, Bureau of Science (1906), January. VOLCANIC TUFF. 393 In addition to the rocks mentioned above, volcanic sediments and pyroclastic tuffs occur quite widely distributed in the Philippines. They, are especially abundant in west central Luzon, extending almost un- brokenly from near Lingayen Gulf to the seacoast of Batangas, practically blanketing or covering nearly all of the massive rocks of this region. Mr. Ickis found volcanic tuff in the Agusan-Pulangui region, interior from Cagayan, Misamis. In the absence of a better stone, in certain places this has been used extensively for building purposes. In Bulacan and along the Pasig River, especially near Guadalupe, this stone is unusually abundant. . Large quantities of it have been quarried and used in the construction of many churches and other buildings and in the walls and fortresses of Manila. It is very workable. Before it is disturbed it is so soft that it can be quarried with an axe, but it hardens rapidly on exposure. Since American occupation the competition of brick and other manu- factured products in Manila has encroached upon the general use of voleanic tuff, nevertheless a large quantity still finds a market. Owing to the extended use of this stone I have made a few experiments to show its real value. The most important laboratory tests which aid one to form an opinion as to the value of a stone are the microscopical, physical, and chemical examinations. Such tests are given below: MICROSCOPICAL EXAMINATION. A microscopic examination of the tuff in the vicinity of Manila shows it to be composed of (1) plagioclase, both decomposed and undecom- posed. There seem to be two generations of feldspar; the one rounded and largely decomposed, the other rather angular and in appearance as if it had come from a greater distance, (2) magnetite, (3) hornblende, (4) quartz grains, (5) the cementing material, which is probably in greater part. volcanic ash, is largely composed of oxide of iron. It might also be mentioned that a certain amount of pumice is nearly always to be found in this tuff. It is undoubtedly andesitic tuff. Silica is the strongest and most desirable bonding material because it is insoluble, free from cleavage and has the same coefficient of expansion as the sand grains which it holds together; however, iron oxides make a fair bond. The latter are more desirable as a cementing material than calcite which is soluble, has very pronounced cleavage, and has a different coefficient of expansion from that of the mineral grains which it holds together, or than clayey matter which is hable to disintegrate. PHYSICAL EXAMINATION. The crushing strength of a few samples of tuff has been determined. The object of such tests in stone is to discover the relation between the crushing strength and the stress to which it would be subjected in a high 394 COX, wall. It is difficult to say what is the effect of long continued pressure upon a stone under this condition, but the majority of architects estimate that ten or twenty times the strength to which it is actually subjected is required. The data are as follows: Tasie I.—Crushing strength of andesitic tuff from near Manila. Ultimate strength per umber Average| Square centimeter. Source. mina- ATCHOS tions S 5 m6 made. | Surface. | Maxi-| Mini- | Aver- | : mum. |} mum.| age. - Sq.em. | Kilos. | Kilos. | Kilos. Manila, quarry unknown______---------_------------ 2 73.3 101 85 | 93 Guadalupe\quarty sess sae eee eee 4 93.5 42 30 | 39.3 Balivag <2. 22-2 ae ee eS 7 54. 205 110 | 151 Majayjay, Laguna quarry _ 3) | see eeenee 194 166 | 181 1 Yo i Ne | PT ls ees 157 141 | 149 2 An old stone which had been used many years in a building. It showed marked sign of decay. Using the average value given for the Majayjay quarry stone and the weight for the wet stone as 1,655 kilograms per square meter (given on p- 399) it can be computed that a wall 110 meters high would be required to equal the compression strength of the stone. Using the safety factor 10, this would be suitable for the erection of structures eleven meters high. Much of the stone from the Meycauayan quarries, Province of Bulacan, which supply stone to Manila is of this class and is fairly durable. The stone formerly quarried in Guadalupe seems to be better than that which is now being taken out. Probably many of the best quarries there and possibly elsewhere have been lost sight of, for recently quantity rather than quality has been sought, i.e., no attention has been paid to sys- tematic testing and that stone which was most accessible has been worked regardless of its grade. Rondelet’s rule that the resistance of a stone to crushing is only three times that offered to traction does not hold for volcanic tuff. The tensile strength of a stone is its ability to withstand a pulling stress. This has been determined for several samples. These determina- tions were made by cutting blocks of the stone of the ordinary size and shape of a cement briquette and breaking them in a standard machine. The rate of the shot in all of these tests was 183 kilos per minute, that adopted by the United States Army in their specifications for cement. Three samples of stone which to my knowledge have been quarried at least two years, taken from various sources in Manila, showed the fol- lowing tensile or breaking strength. VOLCANIC TUFF. 395 TaBLE II.—Tensile strength of stone being used in Manila. e Tensile strength per Tensile strength per Number) square centimeter. square inch. Sam-| termi- ple. | nations Ae: . ae made Max- | Mini- | Aver- | Maxi- | Mini- | Aver- *|mum.|mum./ age. | mum. | mum. | age. Kilos. | Kilos. | Kilos. Lbs. Lbs. Lbs. 1 a 7.38 4.85 5.77 105 69 82 2 4 5.70 5.27 5.48 81 75 78 3 2 8.23 4.64 6.40 117 66 | 91 A sample of stone taken from the Majayajay, Laguna quarry gave the following results : Taste Il1.—Tensile strength of Majayjay stone. Tensile strength per Tensile strength per ae | square centimeter. square inch. termi- | nations Max- | Mini- | Aver- | Maxi- | Mini- | Aver- | 2 | mum. mum.| age. | mum, | mum. | age. | Kilos. | Kilos. | Kilos. | Lbs. | Lbs. | Lbs. 3 | es 6.61 | 8.79 143 93 125 The transverse strength of a stone is the ability to withstand a stress applied at right angles to the length of the block. Its determination is valuable in estimating the thickness of stone required when supported only at the ends or uniformly from end to end. The cracking of stone and brick walls or of single blocks is usually the result of transverse stress due to unequal support throughout their length. The transverse strength of a number of samples was determined in the following manner. Samples having a cross section of approximately 2.5 by 2.5 centimeters, and a length of 18 to 20 centimeters were prepared ; after air drying in my laboratory they were dried at 100° and finally broken in a testing machine. The modulus of rupture was computed from the following formula: __ 2bh? A BY. : We ean R from which Roy W where W=coneentrated load at center in kilograms b=pbreadth in centimeters h=depth in centimeters 1=length in centimeters R=modulus of rupture in kilograms per square centimeter. 396 COX. The results are as follows: Taste 1V.—Transverse strength. Modulus : of rup- Source. Nose Length. Bese, Height. ue pene nape centi- | | meter. | ee __— —— —— ae \'W — Cm. Cm. Cm. Kilos. Kilos. Manila, quarry unknown —__---------- ja 15, 24 2.72 2.89 | 15. 64 15.74 lb 15. 24 2.81 2.97) 21.54] 19.86 Guadalupe quarry *___--------------- { a 15. 24 2.61 2. 83 18.38 | 20.10 b 15. 24 2.77 2.84 27.68 | 28.32 {a 15.24 2.75 2.75 12.70 | 13.96 | b 15, 24 2.78 2.85 15.44} 15.64 e 15, 24 2.61 2.73 12.24] 14.38 | d 15.24 DISS 2.82 13.16 | 14.84 Guadalupe quarry --------------------- ae 15.24 2.72 2.73 | 10.43 | 11.76 | f 15. 24 2.67 2.63 9.30 11.51 g 15, 24 2.77 2.81 10.20} 10.66 | h 15, 24 2.72 2.75 11.33 | 12.59 ea 15. 24 2.59 2.65 9.52} 11.97 Guadalupe quarry » ____-__-----__--__- ja 15. 24 2.79 2. 67 10. 20 bie lb 15.24 2.78 2.85 10,44 | 10.57 Guadalupe quarry ¢ -------=-2-=------- { a 15,24 2.78 2.88 9.75 9.67 | b 15,24 2.74 2.92 10.88 | 10.65 Balivag quarry eee j a 15. 24 2.76 2.67 18.10 21.00 Lb 15.24 2.74 2.68 | 17.20} 20.00 (a 15. 24 2.68 D5) 19.96 | 26.81 a) 15.24 2.60 2.88| 30.66 34.62 CH alos 24 2.62 2.57 24.95 32.96 Majaviay 3 da 15. 24 2.63 2.57 | 20.43 | 26.89 | Majayjay, Laguna quarry__-------_--- 3 Fs e 15.24 2.53 2.84 19.96 | 22.36 f 15.24 2.76 2. 82 32.00! 33.338 g 15.24 2. 82 2. 86 21.77 21.58 h 15.24 2.67 2. 72 27.90 | 32.29 a 15.24 2.61 2.68 20.87 | 25.45 b 15. 24 2.62 2.83 18.17 14.35 @ 15. 24 2. 68 2.83 13. 62 14.78 Majayjay, quarry unknown @-__.—_-__ ia 15.24 2.77 2.70} 24.06) 27.24 | |e 15.24 2, 66 2.73| 26.34 30.37 f 15,24 2.68 2.77 24.06 | 26.75 g 15. 24 2.47 2.78 20.43 | 24.47 a An unused block taken from a pile on the bank of the Pasig River. It had been quarried some time. > Kept in a tightly stoppered bottle so as to retain quarry moisture, dried immediately at 100° when opened. © Kept in a tightly stoppered bottle so as to retain quarry moisture. Not opened until ready to break. It was broken moist with quarry water. 4 An old stone which had been used many years in a building. It showed marked signs of decay. VOLCANIC TUFF. 397 For purposes of comparison I have recalculated the results * of some Wisconsin building stones which are given below: TABLE V.—Transverse strength of Wisconsin building stone. SANDSTONE. | Num- Modulus ber of th of rup- Q sam- | ; Weight | ture per Source. ples Length. | Breadth.) Height. of load. | square aver- centi- aged. meter. | Cm. Cm. Cm. Kilos, Kilos. | Babcock and Smith quarry -____-____- 2 8.89 2.55 2.85 57.8 37.0 Bass Island Brownstone Co. quarry ___| 2 8.89 2.45 2.65 49.4 41.8 C. and N. W. Ry. Co. quarry ___ af 6 10.16 2.58 2. 62 83.0 71.3 Flag River brownstone quarry 2) 17.78 2.78 2.80 25.6 31.3 Grover red sandstone quarry _____-____ 1 7.62 2.70 2.43 35.8 25.5 Grover red sandstone quarry *__ 7.62 2.45 2.45 13.6 10.6 | Prentice Brownstone Co. quarry -_---- 2 12.70 2,72 2.81 41.3 36.5 | | | LIMESTONE. ’ Bridgeport quarry __ : een 1 15. 24 2.69 2.71 71.2 81.9 Story Bros. quarry —___ 2 15. 24 2.41 2.76 120.2 149.7 Marblehead L. and S. Co. quarry ----- 1 15. 24 2.73 2.73 227.3 255.4 Menominee Falls Co. quarry ____------ 1 15.24 2.77 2.62 185.5 22279 Laurie Stone quarry —-------______-___ 1 12.70 2.66 2.59 307.5 327.6 GRANITE. : - Montello Granite Co. quarry —--------= 2 10.16 2.53 2.75 335. 2 266.8 New Hill O’ Fair quarry _--__________- 2 10.16 2.67 2.66 232.2 177.1 4 Sample was wet. The modulus of rupture of the tuff is not high, but it is not necessary that it should be since it is not used as caps and sills in heavy buildings. The necessity of having a high modulus of rupture was obviated in the large buildings, for example the Guadalupe church, by using stone of considerable thickness and by arching the doors and windows. Durability—The durability of a stone depends chiefly upon its ability to withstand the mechanical, physical, and chemical conditions to which it is exposed. The most important durability tests which aid in de- termining this are the specific’ gravity and porosity. Specific gravity—The method employed was to weigh the sample in air and then to weigh it, completely saturated with water, in water at * Buckley, H. R.: Bull. Wis. Geol. & Nat. His. Swr. (Econ. Sr. 2) (1898), 4, 396. 75034——_7 398 COX. a definite temperature and finally to divide the weight in air by the difference and correct for the difference in the specific gravity of water at the given temperature. It is realized that complete saturation is difficult to attain, but I think the method used is satisfactory. A cube of Guadalupe tuff was slowly (several hours were required) immersed in water and finally allowed to remain over night completely covered with water. The results have been computed in two ways, first, on the basis of the sample air-dried in my laboratory for eight months (July to March), and second, on the basis of the sample absolutely freed from interstitial water by heating in an air-bath at 110° for eight hours. The sample retained only 0.7 per cent interstitial water after being air- dried for eight months. The specific gravities (water at 4°=1) are as follows: I. II. 1.895 1.907 Porosity.—To obtain this factor, the weights of the dry and saturated stone were used. The sample was saturated as described above, the surface quickly dried by pressing in filter paper and the weight taken. The difference in the weights was multiplied by the specific gravity 1.907 and the product added to the dry weight. The difference of the dry and saturated weights multiplied by the specific gravity divided by the above sum gives the actual pore space, compared with the volume of the sample tested, as 28 per cent. Temperature changes.—Vhe coefficient of expansion is a most important factor in the disintegration of rocks in many places. Alternate freezing and thawing and extreme heat in case of conflagration are the commonest causes of disintegration in northern countries. Rocks often contain grains of widely different coefficients of expansion and with change of temperature internal strain is produced which aids in the destruction of the rock. A porous rock is somewhat able to accommodate itself to such a strain. This volcanic tuff, as is-true of all very porous stones, can not be used in a cold country because the absorbed water freezes and, breaking the bonds, causes the stone to crumble, but the uniform temperature of the Philippines favors the long life of a soft, porous stone. Water upon freezing at 0° and under one atmosphere pressure expands about one-ninth of its volume. If the water is not allowed to expand it must remain ° Buckley, loc. cit., 69. VOLCANIC TUFF. 399 liquid. James Thomson” calculated that the melting point of ice would be lowered by n 0.0075° for an increase of n atmospheres. W. Thomson ™ confirmed this cal- culation within a difference of 3 per cent by determining the melting point of ice at 8.1 and 16.8 atmospheres. - M. Mousson* kept water liquid at —5° by greatly increasing the pressure and found that at a pressure of about 13,000 atmospheres, ice melted at —18° to —20°. Calculated from the data of Thomson the value for —20° would be but 2,666 atmospheres, about one-fifth the value obtained by Mousson. Assuming the smallest value, it is evident that there is no stone strong enough to resist the strain produced by the freezing of any considerable quantity of water within its pores. In the Philippines the abundance of rain and the high humidity of the air keeps the stone moist during the greater part of the time. The action of the water is markedly twofold, as a solvent for mineral matter and as a solvent or carrier for the gases of the atmosphere. This tuff hardens rapidly after being exposed to the air and this change is undoub- tedly hastened by the presence of moisture. Table III shows an increase of nearly 50 per cent in the tensile strength of Guadalupe stone which had been exposed for some time. Majayjay stone which had been exposed so long that it was considerably decayed, still had a tensile strength about equal to that of the newly quarried stone. Cubic weight.—The cubic weight of this stone as it is taken from the quarry depends upon its specific gravity, porosity, and the water content. With uniform stone the only fluctuating quantity is the water content. Owing to the very large pore space and the heavy rains in the Philippine Islands we may expect this stone to absorb and give up water readily and its eubic weight to vary between the limits for the thoroughly air-dried sample and the saturated one, namely, 1,375 to 1,655 kilos per cubic meter (35.3 cubit feet). The rate at which this stone gives up its water in dry storage is also an important factor in its transportation. Cubes of different sizes were carefully saturated with water and suspended in the laboratory on a 1-centimeters-mesh wire net so that evaporation could take place equally from all surfaces. The tests were begun on March 23, 1908. The data and the results are as follows: 0 Trans. Roy. Soc. Edinburg (1849), 16, 575; Camb. and Dubl. Math. J. (1850) 5, 248; Ann. d. Chim. (1852), 35, 376. u Phil. Mag. (3), (1850), 37, 126. % Pogg. Ann. (1858), 105, 172; Ann. Chim. et Phys. (3), (1859), 56, 252. COX. Taste VI.—The rate at which tuff dries in the air. Water referred to the dry stone in ubes of approximately— e Ferlog : ee } ee | Snead: | Rainfall Time of aération. | ging i : : : ature ity during tion, | 2-2 cubie} 5 cubic | 10 cubic | 20 cubic | during | during | period. | 3 centi- centi- centi- centi- | period. | period. | | meters. | meters. | meters. | meters. | x Z | Hours. | Per cent. | Per cent. | Per cent. | Per cent. °C. Per cent. mm. When saturated __ 0 35. 31 35. 07 35, 52 SER CB) | ee ee Ep OUTS aaa 25 32.05 33. 60 BLO TAD | pee ens 0 64 hours ----__---- 4 24, 35 30. 40 S2 1008 | =aaaeane se 0 223 hours --__----- 16 3.89 19.33 DESY eee 0 263 hours ----____- 4 2.79 NCSA | eee | a b : 0 3034 hours — ™ 4 2.39 15. 25 26.73 30. 97 27.2 75.0 6.1 461 hours ------__- 16 1.99 9.23 23.40 28.75 24.6 88. 4 0.5 54} hours _------__ 8 1. 64 Os 2g | eae | ne 29.2 65. 6 0 WohOUrs esse SHY eke os 2.91 16.13 25, 84 24.1 75,2 0 943 hours _------_- 213 1, 64 1.92 LON583|=asee— = 26.5 66.6 0 1073 hours --_----_ 13 y5| Sono 1.73 8. 64 21.97 27.7 70.0 0 1203 hours -_--_-_- 13 1. 64 1.72 fica /eg| eae 25.1 81.1 0 | 24 27.1 75.7 0.1 160 outs asso ato H 24 } Re os ree { 26.0 87.7| 58.0 216)hourss—se= CY ese A ey 3.30 12.30 26.1 74.3 0 266;hours==seessae 2.54 9.22 27.1 65.3 0 2887 hours ___-____ OA | semen 27.3 68.8 0 3363 hours ~__ 2.16 6.94 27.1 69.4 0 398 hours --------- 1. 5.70 27.9 65.2 0 450 hours ____-____ Io 4.59 28.3 63.6 0 495 hours ~--__-___ il, 4.04 27.5 66.6 0 545 hours -________ 3.56 28.0 65.6 0 591 hours ________- 3.20 27.6 66.3 0 MOM oOurseaeeears 2.59 28.3 68.3 | 0 769 hours ~ 2.30 27.7 61.4 0 1,000 hours 1, 82 28.5 66.4 0 1,070 hours ______- 70 1.80 27.5 77.6 38.2 1, 382 hours _______ {ise | ea Ses eileen 1.94 1.75 28.1 78.0 84.2 After drying at 105° to 110° and cooling ina des- iecators se eee 0.00 0. 00 0:00) | Saas teres | Sane | Seas | eee VOLCANIC TUFF. The foregoing data may be expressed in curves as follows: re Log hours | ae Fie. 1. “LAURA JO ASTI NAIA 401 The aération of the cubes in hours. 402 COX. The irregularities in the curves can be directly accounted for by the variations in the humidity and the temperature of the air from day to day. The curves show this sample of tuff to be in equilibrium with the atmosphere when the latter contains about 1.75 per cent of loosely held water. All of this may be removed by drying at a temperature just above the boiling point of water, but the tuff quickly regains most of it when again exposed to atmospheric conditions. Upon resaturation the cubes absorbed enough water to return to the original saturation value. If there were no other elements influencing the loss of water, the rate of evaporation from blocks should be directly proportional to the surface exposed, i. e., the curve expressing it would be a linear function of the latter; however, several other factors enter. The exposed surface of the cubes used bear to each other the relation 1, 4,16, and 64. It will be noticed (fig. 1) that the cubes reach equilibrium with the atmosphere not in the above time ratio, but in approxianately 36, 108, 324, and 972 hours, respectively, or a ratio of 1,3, 9, and 27. The chief factor causing the differences in these ratios is capillarity and it is constant. Hence the following curve can be plotted, from which the time required for a cube of any size to dry may be directly read: VOLCANIC TOFF. Petes perefere ptf bre pre bape pee ppd paepreder perp peepee pe 2 — 2 2 wate - . e M “BuP9WUED U! aqn> yo ats Fie. 2. 403 ob] ae! Petia ee Po STI fat Tt |oDn] Fen] )enoy [ono jen jon ne one joni ono no ane! [ony )onay jone| jong! jane jane enor Venn erat jon 400 Puepeapey 200 Houra. required for the saturated block to air-dry. eeepe pepe peepee perpen 404 COX. Variability—These few tests show that physically the tuff occurring throughout the Philippines is very variable. When no attention is paid to its selection it may be good or extremely poor and this is why it has fallen into disrepute. With the selection of a good quarry, this stone should be valuable, especially because of its accessibility and the ease with which it is worked. CHEMICAL EXAMINATION. The chemical constitution of several samples of Philippine tuffs and their close relationship to the igneous rocks is shown by the following analyses : TasLe VII.—Analyses of Philippine tuffs and igneous rocks. [Figures give percentages. ] Tuff from— Igneous rock from— Guadalupe. Canlaon Voleano.| Source. ie oa SR Malaqui, Man- Majay-| Taal Aroroy, North South- ila.» . jay. Volca- |Masbate.e rim of| west a. bec no.? er slope | of old | crater. | crater. Silica (SiO,)_------------- 56.84 | 56.55 | 59.27 | 57.26 53. 81 58.14 | 53.69 55. 97 Alumina (AleO3) ~-------- 18.46f, 22.34f/ 17.06 | 16.95 19. 69 17.93 | 18.00 20.35 ic oxi 29 peer 5 1.87 2.16 Ferric Guitele (Fe.O3) 0.75 87 \ 7.55 8.16 5.46 ea 6.26 Ferrous oxide (FeO) —--_-! OE GT ae 2.61 Lime (CaO) ~_------=222-2 4.78 4.74 3.37 3.56 7.73 7.63 8.64 7.92 | Magnesia (MgO) -- eliepel 59) 2.36 1.52 1.10 3.13 2.61 4,59 3.40 Soda (Nas.O) -------------- 4,12 2.38 2.49 1.64 3. 64 4.08 3.07 3.52 | Potash (K,.O)------------- | 2.72) 2.84 3. 63 1. 86 2.19 2.16 1.63 2.48 | Loss on ignition s_________ 6.95 4. 86 6.42 | 7.65 2.13 1.76 1.08 0.38 | | Water (H.O) (below | , 10521109) 1.76 2.51 1.34 1.43 0. 24 0.10 0.31 0.33 Titanic oxide (TiO.) _____ (hn) (4) 0.83} 0.91 (») (») (b) (h) Manganese oxide (MnO)-| trace. |_..----__| trace. O5282): oo a be eo S| ee eee | Total e Es ee 100.48 | 100.44 | 100.70 | 100.14 100. 72 99.87 | 100.12 | 100.61 l a Equal portions of the three samples given in Table I were mixed thoroughly, pulver- ized, air-dried and analyzed. > Sample collected in 1906. ¢ Sample collected in 1908. 4 Average of two very closely agreeing independent samples, probably dolerite. e Average of two very closely agreeing independent samples, probably andesite. t Includes titanic oxide (TiOs). = Mostly water (H20O) above 105-110°. h Included in alumina. CEMENT. I first analyzed the Philippine tuff in the early part of 1907. Its composition is nearly the same as many of the clays and shales used for cement manufacture. Since then I have had hopes that it might be used as a cement material and have expended considerable effort in VOLCANIC FUEL. 405 the attempt to use it for this purpose. In 1887 Le Chatelier ** pro- posed the formula # [(3Ca0)Si0,] + y [(8CaO) Al,O,] for Portland cement, but 8. B. and W. B. Newberry ?* have shown that the formula wv [(8CaO)Si0,] + y [(2CaO) Al,O,] produces a much better cement. Both of these formule are based on a more or less ideal condition of fluxing and there are cases where even the latter gives a percentage of lime in the finished product very much higher than that of the average Portland cement; however, it is valuable to the cement chemists as a limiting formula. Bleininger’® has made a series of investigations from the results of which he concludes that “for the dry ground mix- tures the formula (2.8CaO)Si0,,(2CaO)Al,O, is the safest.” Several cement batches were made up with the sample of tuff from Manila according to the formula for Portland cement, calculated in such a way that after burning the percentage of lime would vary within the limits of good Portland cement. The materials were combined so that the finished product contained : And for each mole- For each molecule cule of Al.Og resp. of SiOz, molecules of Fe,O3, molecules of CaO resp. MgO, as CaO resp. MgO, as follows : follows: 2.8 2 2.6 2 2.4 2 2.2 2 I have not as yet been able to obtain a furnace temperature above 1,350° C and that is too low satisfactorily to burn a cement. All of the above mixtures when burned at this temperature disintegrated spon- taneously on cooling, which is characteristic of the dicalcium silicate (2CaO)Si0,"° and indicates that the heat was not sufficient to fuse the other compounds of silica, alumina, lime and iron oxide which promote the union of silica and lime to form the tricalcium silicate (3CaO)SiO.,, which is the basis of hydraulic activity in Portland cement. In a small crucible over a blast lamp I prepared a cement from the ingredients limestone, clay, and shale which set well, but a sufficient quantity could not be obtained in this way to ascertain the physical constants. With this same method I could not produce a cement from the tuff cement batch, indicating that if cement can be produced from this material a still higher temperature is necessary. Recently Howe” has shown that a cement of good quality can be produced from Panama rhyolite tuff, which is not very unlike that of the Philippine Islands in composition. The Panama materials used were % Ann. des Mines (1887), 11, 345. “J, Soc. Chem. Ind. (1897), 16, 887. “ Bleininger, A. V., The Manufacture of Hydraulic Cements, Geol. Sur. of Ohio (1904) (4), 3, 236. ** Le Chatelier, loc. cit. “ Hoon. Geol. (1907), 2, 655. 406 COX. coral, rhyolite tuff and clay of the following compositions. For purposes of comparison I also give some analyses of Philippine materials. P ename coment Materietneat Analyses of Philippine limestones. Average From A A from mili- Component. From the in- 7 Rayo] | Rul | pasig [Dama] £2 | terior, | From | 122) re Coral.| lite | Clay.| Gun. | clays | Cebu | pay Capiz | Rom- qi hol tuff a ay. coal iar, | near | blon. [Qt poles 5 dalupe. fields Capiz. Maes Nos. 5 and i TFA 6, Batan | 8. Island. | SiQje esses 0.89 | 60.93 | 49.91 | 57.68 | 50.51 0. 36 0.72 0.21 0.10 0.97 Al,03 -_-__---_- 0.32 | 15.86 | 15.48 | 16.60] 20.20 } 0.18 e 51 0.17 \ 0.17 e 56 Fe.Q3 ---------- 0.36 | 5.46 | 10.06} 4.92] 8.08 : 0.31} 0.71 : 0.36 CaO aS 52.62 | 4.02] 6.98 3.28 3.88 | 55.62 | 54.03 | 54.42 | 55.23 53. 86 | 0.38 | 1.79 | 2.27 1.48 2548) Seas 0.99 0. 41 0.45 0.19 tion 43.50 | 10.44 | 12.92 | 10.44 | 12,92] 43.67 | 43.93 | 43.84) 43.80 43.47 * Recalculated to the same loss on ignition as the Panama tuff. » Recalculated to the same loss on ignition as the Panama clay. It will be seen that the Panama coral is a fairly pure calcium carbonate low in magnesia, ideal for the manufacture of cement. The analyses of several samples of Philippine limestones show them to be even purer than the Panama sample. Limestone occurs abundantly throughout the whole Archipelago and is uniformily remarkably pure. From analyses giving the composition of the cement: clinker, the Panama materials were mixed in the following proportions and gave the following data and results for the final product: panic ok Tensile strength of briquettes in kilos per per cent. square penluumeter Ingredients. : 4 4 Neat. Sand (3:1). on 6/5] > wad Eo es i) oS Se ie a ll 3 oe ees $|s Bl@!lsat|oa E chiles || eet “| 2 cel 2 see eso ae reales ee | rect an cay il Spans y =| el & g =| Fee es ee | ee dees |S eS Sar e & 6 S Oe) aan res | ks! | | & a & a a a a ~ & oO © é a ry Ce) | | | Coral-clay______ 2.88 Dy ees 100 | 98 |90.3 |16.7 |44.5 [50.6 (51.7 [57.8 |14.4 pk 2 2. 7 | Leth Coral-tuff ______ Snd05|Saaee| 1 | 100 99 \93.3 |14.8 89.4 |44.7 8 |47.7 |15.9 21. 9 (27.6 30.0 Coral-clay tuff -}_____|_____]_____ 100 95 |88.6 |20.3 i 1 (51.6 (a7. (6) | eee 12.2 21.2 lo9, 4 eee | \ | "One kilogram per square centimeter —14.22 pounds per square inch. It is interesting in the above table to note that the coral-tuff cement gives the strongest sand mortar. The laboratory is now preparing to install a small rotary cement kiln and when this is completed further experiments will be carried on in this direction. ILLUSTRATIONS. i Page. Fie. 1. (In text.) Curve showing the rate at which blocks of tuff give up WIETIP WEY i GHZ. UOTE repens raster Sana tet Oe ase oreo nea ROR 401 2. (In text.) Curve from which the time required for a saturated cube of tuff of any size to dry may be directly read..........2.---20.------2----- 403 EDITORIAL. THE EFFECT OF LITSEA CHINENSIS ON THE HARDENING OF LIME MORTAR. The leaves of the plant known to the Tagalogs as puso-puso (Litsea chinensis Lam.) have long been in favor with native masons for use in making mortars. Their present method of treating the leaves of the puso-puso for this purpose is to beat the fresh leaves into a pulp, adding water until the mixture is a pale green. ‘his is rather sticky and becomes more so as it ferments. After it is thus prepared it is allowed to stand and ferment for from twelve to eighteen hours, when it is used with native lime. Undoubtedly these leaves have been employed for this purpose for a long time as Blanco’ tells us that puso-puso is well known in the Philippines, the leaves when infused in water for six or eight hours forming a mucilaginous substance, they being cut into small pieces with a knife and triturated very well before being put into the water. When the decoction is mixed with lime and sand it is stated that a very strong mortar is made which is said to be almost impermeable to rain. The puso-puso is a tree 7 to 9 meters high with a very hard wood. It is a species widely distributed from southern (tropical) Asia to Malaya. Governor Sandiko, of Bulacan Province, reports that the native masons make their cement of lime 30 parts, sand 60 and melaza 1 part; the water used being the puso-puso juice prepared as above described. Mr. Warner, of the Bureau of Supply, using the puso-puso decoction and comparing its results with those obtained with lime and sand mixed with water alone obtained the following table. In all the tests 184 per cent of water, respectively, puso-puso solutions were used. The tensile strengths are as follows: Time of set. Average. One Two Three One Two Three | month. | months. | months. | month. | months. | months. | | | 38 , © 88 37 62 90 GTM e-San d= WA be re é 0. 88} | 37 60 87 vas SOF 35 61 88 | | 48 53 7 | | 50 55 82 | Lime-sand-melaza-puso-puso juice __ 50 56 81 492 563 822 49 61 90 | | 50 56 81 | | * Blanco, Flora de Filipinas, 2d ed. (1845), 566. 409 410 EDITORIAL. The above experiments show that puso-puso and (melaza) increases the initial rapidity of hardening, but the indications are that in the final set the tensile strength is the same or somewhat less than when water is used alone. In recent years tests of the influence of a great many indifferent substances have been made on the hardening of mortar with small positive or negative results. Parsons and Porter* experimenting with Portland cement found that the addition of sugar or molasses delayed the setting of the mortar, the retardation being greater when molasses was used, but when certain pro- portions were not exceeded, the strength of the mixture was slightly greater than that of the pure cement. Sugar apparently has no chemical action on mortars. The variation of the binding power is due more to mechanical causes and probably favors or retards the chemical reactions involved. Rohland* concludes that the hardening of cement is not due to the formation of chemical compounds, but to a specific action of colloids, and perhaps the same might be said of lime mortar. I believe that in the final equilibrium few, if any, foreign bodies have a positive influence. The above-described use of puso-puso juice appears to give no explana- tion of the extreme hardness of some of the old-mortars. In general the hardening of lime mortar depends on the chemical reaction Ca(OH ),--CO,—>CaCO,-+-H,0. This reaction really exhibits two phases, namely, 1. Ca(OH),—>CaO-+H,0 and 2. Ca0-+CO,—>CaC0O,. The velocity of the reaction is extremely slow, requiring months, years, and in some cases even hundreds of years to complete it. The difficulty with which gases penetrate solid walls is shown by the fact that reénforcing steel embedded in concrete will remain free from rust. Recently in tearing down a one-story building erected in 1902 at New Brighton, Staten Island, all steel reénforcements were found in perfect preservation excepting in few cases where they were allowed to come closer than three-fourths inch to the surface.* Uncombined calcium hydroxide has been found in mortar inclosed between very compact stones after they had been in place three hundred years. In the light of these facts the explanation of the superior quality of the old lime mortars found in various places is probably that the porous stone so ? Dingler’s Polyt. Journ. (1889), 271, 268; J. Soc. Chem. Ind. (1889), 8, 545. *Ztschr. f. Elektrochem. (1907), 13, 11. “Turner, H. C. Eng. News (1908), 59, 75; Anon, Jron Age (1908). 81, 348. EDITORIAL. 4\1 generally used renders it possible for the carbon dioxide of the air to enter the interior parts of the massive masonry and so to hasten the equilibrium of the hardening reaction. AtvIn J. Cox. PROPRIETARY MEDICINES IN THE ORIENT. The drugs inspection under the Philippine “Food and Drugs Act of May 18, 1907” is revealing many illegalities and peculiarities in the proprietary medicines of the Orient. The Chinese medicines offered at the various ports for entry into the Philippine Islands are for the most part composed of pastes and powder manufactured from the herbs and animal products. Sea horses are caught in large number, dried, powdered, and compounded into medi- cines. The most abundant species is Gasterotokeus biaculeatus (Bloch) ; others are Hippocampus kuda Bleeker and H. aterrimus Jordan and Snyder. Other medicines are supposed to be manufactured from various parts of tortoises and stags. The advertising literature, in Chinese, accompanying an invoice of pills states, “These pills are prepared from the best ginseng and deer’s horn obtainable in this country,’ and recommends their use for all pains having neuralgic origin, rheumatism, liver and intestinal diseases, nervous exhaustion, overstudy and sleeplessness. Another variety of pills was labeled with the modest statement, “These pills are used as a tonic in the treatment of all diseases occurring among both sexes and all ages. For the healthy they are especially useful as a preventive in con- tracting disease.” Then follows an enumeration of the ailments for which the pills are especially adapted, among which may be mentioned, general debility from overwork, alcohol, sleeplessness, rheumatic pains, toothache, headache, pain in the lumbar regions, heartburn, nausea, vomiting, loss of appetite, swelling of glands, and cedema of legs, face or abdomen. Many Chinese “medicinal beverages” are offered for entry at this port. One, labeled as a cure for rheumatism and dropsy and a tonic for lungs and liver, owed its effect to an alcoholic content of 50 per cent. The usual varieties of Chinese pills containing morphine and opium are constantly met masquerading under various names and false state- ments as to their character and composition. One shipment of small, black pills, advertised and labeled as a cure for the opium habit was found to contain a large percentage of morphine. While a large proportion of the Chinese medicines are of the proprie- tary class the patent medicine business of the Orient is by no means EDITORIAL. — 412 confined to the Chinese. Some establishments in this part of the world under the control of foreigners, place upon the market preparations which are in many cases more reprehensible in their composition and in the character of their advertising than the medicines manufactured by the Chinese. Proprietary preparations from Japan, United States, France, Spain, Holland, Germany, and some other countries are also to be found upon the local markets. Among those which have been found to be in violation of the “Food and Drugs Act” the following are typical examples. A Spanish stomach tonic and a Japanese “injection” were both found to contain cocaine. A Spanish pectoral paste contained heroin and a “cholera elixir’ was composed almost entirely of morphine and chloral hydrate. Japanese “brain pills’ were found to owe their effect to the usual quantity of acetanilid. A widely advertised “dysentery cure” much used among the foreign element in the Orient was found to contain calcium carbonate 90.7, calcium phosphate 3.7, organic matter 3.6 and water 2.0 per cent, and is reputed to be either ground cuttlefish bone or ground oyster shells. Two bottles of liquid, shipped by mail, and pur- porting to be a physician’s prescription, were found to be concentrated aqueous solutions of morphine sulphate. Another preparation which “works miracles with every one that makes use of it, and the Grace of the Omnipotent God is experienced in it to admiration” is recommended in the accompanying advertising matter as a cure for stones in the bladder, bruises in the hands or feet, black and blue spots, thickness of blood, all kinds of fevers, asthma, liver troubles, hysterical pains, dropsy, the French disease, worms, palpitation of the heart, headache, burns, colic, and if put in the eyes “it will make you so strong sighted that you need not use spectacles until the age of 70 or 80 years, thus preserving the sight.” Analysis shows the substance to be a fish liver oil. The Narcotic Drug Law (Act No. 1761 of the Philippine Commission, October 10, 1907) so restricts the use of opium, cocaine, alpha and beta eucaine that many proprietary medicines and preparations are denied entry to the Philippine Islands. f H. D. Gripss. Vou. WI DECEMBER, 1908 No. 6 THE PHILIPPINE ~ JOURNAL OF SCIENCE EDITED BY PAUL C. FREER, M. D., Pu. D. WITH THE COOPERATION OF MERTON L. MILLER, Pu. D.; GEORGE F. RICHMOND, M. S$. W. D. SMITH, Pu. D.; A. J. COX, Pu. D. RAYMOND F. BACON, Pu. D.; CHARLES S. BANKS, M.S. , H. D. GIBBS. B. S.; R. C. MCGREGOR, A. B. PUBLISHED BY THE BUREAU OF SCIENCE OF THE GOVERNMENT OF THE PHILIPPINE ISLANDS A. GENERAL SCIENCE MANILA BUREAU OF PRINTING 1908, PREVIOUS PUBLICATIONS OF THE BUREAU OF GOVERNMENT LABORATORIES, 1No. 1, 1902, Biological Laboratory.—Preliminary Report of the Appearance in the Philippine Islands of a Disease Clinically Resembling Glanders. By R. P. Strong, M. D. No. 2; 1902, Chemical Laboratory.—The Preparation of Benzoyl-Acetyl Peroxide and Its Use as an Intestinal Antiseptic in Cholera and Dysentery. Preliminary Notes. By Paul C. Freer, M. D., Ph. D. 1No. 8, 1903, Biological Laboratory.—A Preliminary Report on Trypanosomiasis of Horses in the Philippine Islands. By W. EH. Musgrave, M. D., and Norman EH. Williamson. 1No. 4, 1903, Serum Laboratory.——Preliminary Report on the Study of Rinderpest of Cattle and Carabaos in the Philippine Islands. By James W. Jobling, M. D. 1No. 5, 1903, Biological Laboratory.—Trypanosoma and Trypanosomiasis, with Special poterence to Surra in the Philippine Islands. By W. E. Musgaye, M. D., and Moses . Clegg. u 1No. 6, 1903.—New and Noteworthy Plants, I. The American Hlement in the Philip- pine Flora. By Elmer D. Merrill, Botanist. (Issued January 20, 1904.) 1No. 7, 1908, Chemical Laboratory.—The Gutta Percha and Rubber of the Philippine Islands. _By Penoyer L. Sherman, jr., Ph. D. 1No. 8, 1903.—A Dictionary of the Plant Names of the Philippine Islands. By Elmer D. Merrill, Botanist. 1No. 9, 1908, Biological and Serum Laboratories—A Report on Hemorrhagic Septi- cemia in Animals in~the Philippine Islands. By Paul G. Woolley, M. D., and J. W. Jobling, M. D. 1 No. 10, 1903, Biological Laboratory.—Two Cases of a Peculiar Form of Hand Infection (Due to an Organism Resembling the Koch-Weeks Bacillus). By John R. MeDill, M. D., and Wm. B. Wherry, M. D. 1No. 11, 1903, Biological Laboratory.— Entomological Division, Bulletin No. 1: Prelimi- nary Bulletin on Insects of the Cacao. (Prepared Especially for the Benefit of Farmers.) By Charles 8. Banks, Entomologist. 1 No. 12, 1908, Biological Laboratory.—Report on Some Pulmonary Lesions Produced by the Bacillus of Hemorrhagic Septicemia of Carabaos. By Paul G. Woolley, M.D. No. 13, 1904, Biological Laboratory.—A Fatal Infection by a Hitherto Undescribed Chromogenic Bacterium: Bacillus Aureus Fetidus. By Maximilian Herzog, M. D. 1No. 14, 1904.—Serum Laboratory: Texas Fever in the Philippine Islands and the Far East. By J. W. Jobling, M. D., and Paul G. Woolley, M. D. Biological Laboratory: Entomological Division, Bulletin No. 2: The Australian Tick (Boophilus Australis Fuller) in the Philippine Islands. By Charles §. Banks, Entomologist. No. 15, 1904, Biological and Serwm_Laboratories—Report on Bacillus Violaceus Ma- nile: A Pathogenic Micro-Organism. By Paul G. Woolley, M. D. 1 No. 16, 1904, Biological Laboratory.—Protective Inoculation Against Asiatic Cholera: An Experimental Study. By Richard P. Strong, M. D. No. 17, 1904.—New or Noteworthy Philippine Plants, Il. By Elmer D. Merrill, Botanist 1No. 18, 1904, Biological Laboratory.—I. Amebas: Their Cultivation and Etiologic Significance. By W. HE. Musgrave, M. D., and Moses T. Clegg. II. The Treatment of Intestinal Amebiasis (Amebic Dysentery) in the Tropics. By W. E. Musgrave, M. D. No. 19, 1904, Biological Laboratory.—Some Observations on the Biology of the Cholera Spirillum. By W. B. Wherry, M. D. : No. 20, 1904.— Biological Laboratory: 1. Does Latent or Dormant Plague Exist Where the Disease is Hndemic? By Maximilian Herzog, M. D., and Charles B. Hare. Serum Laboratory: Il. Broncho-Pneumonia of Cattle: Its Association with B. Bovisepticus. By Paul G. Woolley, M. D., and Walter Sorrell, D. V. S. III. Pinto (Pano Blanco). By Paul G. Woolley, M. D. Chemical Laboratory: IV. Notes on Analysis of the Water from the Manila Water Supply. By Charles L. Bliss, M.S. Serum Laboratory: V. Frambesia= Its Occurrence in Natives in the Philippine Islands. By Paul G. Woolley, M. D. No. 21, 1904, Biological Laboratory—Some Questions Relating to the Virulence of Micro-Organisms with Particular Reference to Their Immunizing Powers. By Richard P. Strong, M. D. = No. 22, 1904, Bureau of Government Laboratories.—I. A Description of the New Build- ings of the Bureau of Government Laboratories. By Paul C. Freer, M. D., Ph. D. If, A watalorue of the Library of the Bureau of Government Laboratories. By Mary Polk, ibrarian. ‘ 1No. 28, 1904, Biological Laboratory—Plague: Bacteriology, Morbid Anatomy, and Histopathology (Including a Consideration of Insects as Plague Carriers). By Maximilian Herzog, M. D. No. 24, 1904, Biological Laboratory.—Glanders: Its Diagnosis and Prevention (Together with a Report on Two Cases of Human Glanders Occurring in Manila and Some Notes on the Bacteriology and Polymorphism of Bacterium Mallei). By William B. Wherry, M. D. No. 25, 1904.2—Birds from the Islands of Romblon, Sibuyan, and Cresta de Gallo. By Richard C. McGregor. No. 26, 1904, Biological Laboratory—The Clinical and Pathological Significance of Balantidium Coli. By Richard P. Strong, M. D. No. 27, 1904—A Review of the Identification of the Species Described in Blanco’s Flora de Filipinas. By Elmer D, Merrill, Botanist. No. 28, 1904.—I. The Polypodiaceze of the Philippine Islands. JI. Edible Philippine Fungi. By Edwin B. Copeland, Ph. D. No. 29, 1904.—I. New or Notewsrthy Philippine Plants, III. II. The Source of Manila Elemi. By Elmer D. Merrill, Botanist. x No. 80, 1905, Chemical Laboratory.—I. Autocalytic Decomposition of Silver Oxide. II. Hydration in Solution. By Gilbert N. Lewis, Ph. D. No. 81, 1905, Biological Laboratory.—I. Notes on a Case of Hematochyluria (Together with Some Observations on the Morphology of the Embryo Nematode, Filaria Nocturna)- By William B. Wherry, M. D., and John R. McDill, M. D., Manila, P. I. II. A Search Into the Nitrate and Nitrite Content of Witte’s ‘‘Peptone,” with Special Reference to Its poienge onthe Demonstration of the Indol and Cholera-Red Reactions. By William B. erry, M. D. 1 Out of print. 2 The first four bulletins in the ornithological series were published by the Ethnological Survey under the title “Bulletins of the Philippine Museum.’’ Later ornithological publications of the Government appeared as publications of the Bureau of Government Laboratories. (Concluded on third page of cover.) 6. III, No. Scr., Vou. JOURN, [PuHtI.. IGorors. ] THE BENGUET BEAN: PLATE THE RAIEIPRINE JOURNAL OF SCIENCE A. GENERAL SCIENCE WO, OU DECEMBER, 1908 No. 6 THE BENGUET IGOROTS. A SOMATOLOGIC STUDY OF THE ~ LIVE FOLK OF BENGUET AND LEPANTO-BONTOC, By Rosert BENNETT BEAN. (Prom the Anatomical Laboratory, Philippine Medical School, Manila, P. I.) INTRODUCTION. During the intersessional vacation of the Medical School in the year 1908, I spent two months at Baguio, the capital of Benguet Province, in studying the physical characters of the natives. No casual observer would expect to find white people inside of brown skins, but I found European types among the Igorots. (Plate 1.) Trips were made to Atok, Tu- blay, and Capangan with parties under the direction of William H. Pack, governor of Benguet Province, to whom credit and thanks are tendered for his kindly cooperation in the work and his assistance in establishing the good will of the natives. Dean C. Worcester, Secretary of the Interior, also has my sincere gratitude for enabling me to carry through the work and for his many personal favors during its progress. I made excursions to points near Baguio from time to time, and obtained a few additional measurements at the Benguet Sanitarium and among the camps of laborers located in the vicinity of the town. Benguet Province is situated in the central part of northern Luzon ; Baguio, the capital, being somewhat less than 300 kilometers due north of Manila and about 30 Inlometers east of the seacoast. The mountains of Benguet form a part of the Cordillera Central del Norte of the island, the most inaccessible portions of which lie in the north of the province and in Lepanto-Bontoc. Baguio les at an altitude of 1,500 meters above sea level. It has a temperate climate and is located among pine hills on an irregular plateau southwest of the center of the province at 78322 413 414 BBAN. the terminus of one of the most remarkable highways of the world. The latter is to a large extent carved out of solid rock and in many places the deep cafion of the Bued River is crossed by suspension bridges. The plateau on which Baguio is located rises northward along the west of the province in the form of a group of rugged mountains intersected by small streams that cut their way through narrow gorges to Lingayen Gulf and the China Sea, passing through Pangasinan as well as Union Provinces, the latter inhabited by Ilocanos, one of the most thrifty and energetic people of the Philippines, and great colonizers. The Ilocanos form the littoral population of the west coast of northern Luzon and have penetrated the mountains to some extent. The eastern part of the province from its extreme northern end to its southern limit is drained by the tributaries of the Agno River, beyond which are moun- tains separating it from the Province of Nueva Viscaya. The Province of Benguet is thus divided into mountain.and valley, or highland and lowland. The entire province is practically inaccessible, except over the Benguet Road, over the Naguilan trail from San Fernando, Union, or the trail from Aringay, Province of Union. The rivers that pass out of the province are filled with water during part of the year; their beds are rough, the sides precipitous and the mountains steep and rugged, so that both mountains and rivers form very difficult ways of entry. The present governor is rapidly constructing trails in the mountains with a grade of from 3 to 5 per cent, the most audacious of these is nearing completion and will connect Benguet with the Province of Lepanto- Bontoe. The inhabitants of this isolated region could have arrived only by crossing high and rugged mountains, or by picking their way along the beds of the rivers during the dry season. Whether they came of their own accord or were forced from the lowlands by other peoples may never be known. I believe the Igorots pushed into the mountains as bold pioneers in much the same way that the Puritan, the Scotch-Irish and the Cavalier crossed the Appalachians and settled the western part of the United States. They probably exterminated or absorbed any pre- vious inhabitants and have built for themselves enduring monuments in their rock-ribbed and terraced rice paddies, and in the rock shelters for their dead. Their muscular development is phenomenal (19,70,) and would put to shame the best American athletes. Their laws and customs are founded on justice and equity, and “‘an eye for an eye and a tooth for a tooth” is often carried out to the letter. Civilization has not yet— greatly affected the Igorots and they are being protected from its evil influences as carefully as sedulous officials can protect them. They are one of the few uncivilized communities that civilization has touched yet not defiled. The people of Atoc, in the western part of the province are a rep- “resentative group of Igorots. Atoc is a bold point that juts out from THE BENGUET IGOROTS. 415 the surrounding mountains at an altitude of 2,000 meters above sea level, and its precipitous sides furnish an almost impassible barrier against attack. The inhabitants of this region were the last of the Benguet Igorots to come under the jurisdiction of the United States of America, and it was only by superior force of arms that they finally submitted. They live on their rocky fortress, work the paddy fields of the valleys below, and return to their stronghold at night. They are a self-reliant and progressive people, with sound judgment and wise deliberation in their councils. The administration of their affairs is in their own hands under the guidance of the governor of the province. Their chief baknon (old man) has already roofed his house with galvanized iron for pro- tection from the tremendous downpours of rain which are so frequent in this region, and others are following his example. Men and women are on practically an equal footing. The men work away from home for means to provide food, shelter, and draft animals (for working the paddy fields), and when at home the men care for the children. The women work at home raising the small crops (ca- motes, coftee, etc.), prepare the food, and assist the men in the transporta- tion of surplus products to distant markets over steep mountain trails, acting with the men as common carriers. The women also have a voice in the councils and often exercise a controlling influence. The life of the Igorots is an existence of ideal sexual equality in many respects, and civilized nations might profit by their example, for they impressed me as a remarkably contented and cheerful people. However, the purpose of this article is not to present the moral qualities, but the physical characters of the Igorots. Observations and measurements were made of 104 adult males (16+ years), 10 adult females, and 30 boys between the-ages of 5 and 15 years inclusive. [ also measured a number of Japanese, Chinese, Ilocanos, and Tagalogs, but these data will be reserved to be presented in later papers on the various Filipino peoples. Four groups will be considered in detail with each observation or measurement. The nativity of the four groups is: Lepanto-Bontoe [15], mountains of western Benguet [73], the Agno River valley [30], and Baguio and vicinity [27]. Fourteen of the Le- panto-Bontoe Igorots are adult males, and 1 is a boy. Forty-six of the mountain Igorots are adult males, 10 are adult females, and the re- mainder are boys. Twenty-two from the valleys and 22 from Baguio and vicinity are adult males, the remainder boys. (Table I.) METHODS EMPLOYED. The body parts are measured from the ground up by means of a graduated vertical rod with a sliding horizontal pointer. Other meas- urements are made with sliding calipers (compass d’appaisseur-Colin). Lead electric fuse wire is used in taking head outlines, a hinged brass bar is employed to measure the facial index, and general descriptions 416 BEAN. are given of hair, brows, eyes, ears, nose, and skin, with occasional sketches. Measurements and observations are made on the naked body of each individual, except the women, the usual breech clout not interfering at all. Some difficulty is experienced in obtaining one measurement, that of the superior extremity of the great trochanter of the femur, due to the solidity and rigidity of the hip muscles as the individual stands in the erect posture. My measurements of the Igorots follow the personal instructions which I received from Professor Manouvrier in Paris during the sum- mer of 1906, and I wish at this time to express my gratitude for the painstaking care exercised by him through the course of my training. The methods employed have been used by me during the past two years, and as they will apply to future work on the natives of the Philippine Islands, a brief résumé of the most important is inserted here. The individual should stand in the position of a soldier(67). Projections are then made with the anthropometer as follows, using the level of the soles of the feet as the base: HEIGHTS. 1. Body height—Allow the beam to fall with a click on the top of the head (vertex). 2. Ear height—The beam should point into the external auditory meatus (meatus acusticus externus). 3. Chin height.—The point at lower edge of mandibular-symphysis (protube- rantia mentalis). 4. Sternwm.—Press down firmly in the suprasternal notch (incisura jugularis). 5. Umbilicus.—The level. of its middle. 6. Pubis—The superior border of pubic hair. (I use the actual level of the pubic spine [twberculum pubicum]). 7. Acromion.—The level of its outer tip. 8. The elbow (cubitus).—The level of the joint furrow in the flesh, at the head of the radius (capitulum radii). 9. Wrist (carpus).—The level of the lower extremity of the styloid process (processus styloidus) of the radius. 10. Tip of middle finger (digitus medius) with hand extended. ll. Trochanter.—Press extremely hard on the upper end of the femur (tro- chanter major) . 12. Knee (genw)—The line in the rear on the skin passing exactly through the joint at the upper outer end of the tibia (condylus lateralis). BREADTHS. (Made with triple elbow calipers.) 1. Shoulder.—Press hard on the outer tips of the acromion processes. 2. Hip—The outer lips of the iliac crests (crista iliaca) . 3. Thigh—The outer part of the trochanter (trochanter major). 4. Pelvis——From the anterior superior edge of the symphysis to the diamond- shaped depression in the back, over the lumbar region. THE BENGUET IGOROTS. ALT The masterful work of Rudolph Martin (“Die Inlandstamme der Malayischen Halbinsel” (25)) is freely utilized in the course of the present study because of the complete presentation in tabular form of the results of recent investigators regarding the people associated with eastern Asia. The present article is divided into eight parts the last three of which form a summary; these parts are as follows: I, Stature; II, Body Parts; III, Head Form; IV, Physiognomy; V, Descriptive Characters; VI, Somatologic Race Types; VII, Three Selected Types; VIII, Supple- mentary Theory of Heredity. I. STATURE. The Igorots are a people of small stature (below 160 centimeters) although many individuals are above the average and some are tall. (Table III.) The average or mean height of 104 adult males and 10 adult females is 154.0 and 146.7 centimeters, respectively. Further analysis reveals the fact that these groups are not homogeneous. Only 60 per cent of the adult males are between the height of 150 and 162 centimeters, a wide range for so small a number of individuals. The mode or height of greatest frequency (hence the fashion) is 150 centimeters, although there is only one less individual at 152, 154, and 156 centimeters respectively. The median (which has an equal number of individuals above and below it) is 153 centimeters. The minimum is 141 centimeters, and the maximum is 170 centimeters. There is an even distribution of individuals between 148 and 153 centimeters; there are 25 above 158 and 11 below 148 centimeters. (Table II.) A curve constructed from the number of individuals at the various heights represented by ordinates and abscisse would not be a normal Gaussian curve, but would be platykurtic (flat-topped(74) ) (45) with a tendency toward tallness, indicating great variability, and more than one type of man. The mean height of 14 Bontoe Igorots is 158.6 centimeters. One is only 148 centimeters high, one about 156, and 7 are about 164 centimeters. The three groups are significant when considered in connection with similar ones from the highland and lowland regions. The mean height of 46 adult male Igorots of the highland region is 154.9 centi- meters, the minimum is 142, and the maximum is 170. The height is less than that of the Bontoe Igorots, and the variability is greater, but the height is more than that of the lowland Igorots. The mean height of 22 adult males from the valley is 153.6 centimeters. The smallest is from Trinidad in the open country, and the two tallest are from Buguias, which is in the northern end of the province, near Lepanto-Bontoc. The mean height of the 5 men from Buguias is 155.2 centimeters, while that of the 5 from Trinidad is only 152 centimeters. This would suggest that there is an element of small people at Trinidad. The 5 men from Baguio and the 5 from Kabayan have a mean height of 156.8 and 151.2 centimeters respectively, which would indicate the same for Kabayan. However, five individuals are not enough from which to determine a mean height, although the measurements do indicate the characteristics of a part of the population. The mean height of 22 adult males from Baguio and vicinity is 149.1 centi- 418 BEAN. meters. Their nativity was not ascertained for lack of an interpreter, but many of these probably come from Trinidad. For this reason the group is not characteristic for the whole province, but for Trinidad, Baguio and vicinity. The individuals of each of the three groups, Lepanto-Bontoc, Highland, and Lowland, may be divided into those of small, those of intermediate, and those of great height. The height of the small individuals varies around 148 centimeters (cf. Negrito); that of the large individuals around 165 centimeters (cf. European) ; and the height of the greatest number of individuals is about 154 centimeters (cf. Malay). The people of Atoe are slightly above and the people of Baguio and vicinity are slightly below the figures given, but the three groups are definitely represented there as elsewhere. The conclusion from the examination of the height alone is that at least three groups of people make up the Igorot population. (Table II.) The mean height varies directly with the altitude, but probably this variation is not due to the effects of mountain or river, but to the difference in type of the individuals making up the population. The accessible parts have been influenced by infusions of blood from outside of the Province of Benguet, whereas the inhabitants of the inaccessible regions are more like the original type. However, it is possible that outsiders of a bold and daring nature penetrated to the most inaccessible regions, and these may have been tall individuals who increased the average height of the community by their presence and by their progeny. The mean height of the Igorots is 3.6 centimeters greater than that given by Martin(26) for the inhabitants of the Malay Peninsula. It is also greater than that of the Veddahs of Ceylon (Sarasin), but it is less than that of the Annamites(14), the Japanese, the Koreans, the Javanese and various other peoples in the region adjoining eastern Asia. Sexual differences in height can not be fairly stated because so few women were measured, but it may be of interest to note them. The mean height of the female Igorots is 146.7 centimeters, the minimum 135 centimeters and the maximum 154 centimeters. The mode is 146 centimeters and the median 145 centimeters. The female height throughout the world is 7 per cent less than the male, or in other words the female is 93 per cent of the male height(27). Therefore, since the mean height of the Igorot women should be only 144.1 centimeters they are proportionately taller than the men. STATURE AND RACE. (50) Although local conditions acting on the same people for many thou- sand years may effect a change in stature, yet it remains true that stature is a potent factor in race differentiation. Food and nutrition play a part.in determining this characteristic, and artificial selection is at work in modern social life, tall individuals being selected in marriage because goodly stature in youth implies a bountiful store of vitality. Occupa- THE BENGUET IGOROTS. 419 tion and habitat may influence it in the individual, but this is not transmitted to the offspring. City life reduces stature, but attracts tall men, so that the one balances the-other. The tall, hardy pioneer survives in the mountain, but poor nutriment causes a decrease in height; hence a similar balance is found there. However, these influences act only on the individual and if they become hereditary it must be after count- less generations. Racial differences in stature are characteristic and persistent. The Malays are everywhere inclined to be short, and the Polynesians are inclined to tallness. The Scotch are the tallest people of Europe, the southern Italians are almost dwarfs; the first live in the mountains and the latter inhabit the coast. The Adriatic has a body of very tall people along its northern borders, but the mountains of middle Europe are inhabited by short individuals. The Teutonic people have retained their height wherever they have gone. The inhabitants along the shores of Brittany, which were ravaged so fiercely by these northern barbarians, are taller than the people of the interior. The valleys south of Germany (Tyrol) have been infiltrated by the tall invader, leaving the short man in the mountains. Many other instances could be cited to prove that stature is incident to race, but as the weight of evidence is in favor of this, the burden of proof rests with the opposition; as well argue that long heads are due to mountain height, because long-headed Igorots are found high up in the mountains, as to say that short stature is due to high altitude, or vice versd. However, it may be that stature, like so many other charac- ters, becomes altered by environment in the life of a single individual, but the altered condition is not transmitted, until, through countless generations in the same environment, the altered character becomes fixed and inheritable. The stature of the Igorots then, is probably a racial character, and not a local condition. STATURE AND AGE. The stature of the boys as contrasted with that of the adult males, and the relation of growth to age deserve consideration. The ages which I have given are not exact in every instance, because age is determined by the number of rice harvests since the birth of the individual, but as the rice harvest is annual, this method of record is fairly accurate. The individuals are arranged in small groups from the age of 5 to 20 years, and in larger groups above this age. (Table IV.) The mean stature increases about 5 centimeters per year up to the age of .16, when the adult height is apparently reached, although a slightly greater height is found between 20 and 30. The height at 18 and 19 is less than that at 16 and 17, but the small number of individuals at 18 and 19 may account for this. For the same reason, the great height of 5 men above the age of 50 years, and that of the boys from 5 to 10 years of age, are not fair estimates. 420 BEAN. The two most significant features of the relation of height to age are the apparently early maturity of the Igorots and their aquisition of maximum height at an earlier age than Europeans. This agrees with the conclusions of Martin(28), regarding the inhabitants of the Malay Peninsula. Hastings(17) has presented the average height for each age of 8,245 typical male American school children, his figures compared with the height of the Igorots shows a difference of about 10 centimeters in favor of the American children at each age up to 17, and a further increase of about 10 centimeters to the age of 19, when the American boy is 20 centimeters taller than the Igorot. The actual height as well as the growth of the American children conforms well with Topinard’s deductions from measurements of 1,104,841 Europeans(61). The age in the latter instance is carried beyond 30 and the greatest height is found to be between 30 and 40 years. There is an annual increment in height up to 35 years(60). This increment decreases during the pe- riod of from 1 to 5 years, makes a sudden slight increase at 6, remains stationary from 7 to 10, increases progressively from 11 to 16, decreases suddenly at 17 and slowly thereafter to the age of 35, when the increase in height ceases. The growth of the Igorots is similar to this. The stature increases steadily from 10 to 17 years, there is a decrease to the age of 20, then an increase to the maximum between 20 and 30. After 30 the height decreases slightly to the age of 50. The relative height increment of the Igorot boy is not unlike that of the European girl, because the annual increment decreases in both from the age of 13 to 19 (69). DISCUSSION OF STATURE. In conclusion it may be said that the growth of the Igorot is similar to that of the European, but that it is more rapid. The Igorot male is as well developed at the age of 16 years, as the European at 18. The maximum height of the Igorot is reached between 20 and 30, that of the European ten years later. The relative growth of the Igorot boy is in- termediate between that of the European girl and the European boy. The height of the different groups of Igorots(73) varies directly with the altitude and inaccessibility of their location, but the rate of growth, the time of maturity and the actual height are probably characteristic of the stock and not due to environment. The stature of the adult male Igorots is represented by a curve which is seen to be irregular. (Fig. 1.) With only 104 individuals some irregularities in the curve might be expected, which would be smoothed “if 1,000 had been measured, but evidence indicates that irregularities in a curve of 100 individuals mean a diversity in type due to previous mix- * European is used in the sense of the white or Caucasian. 21 “1 Ht ‘ousou ey — a ‘oVINN =O seoiseu o[vul ——— —o1nsy puev o9[vos [[euUg ‘OuseU JSBON BouINH—gq "sjO103] a[eVuley — =oinsy puv o[vos o81v7 ‘OylIZeN ‘ueuMYysng ‘j0}0e7]}0H = V *sjO103] 9] BUI "Sso0130U o[VUL 94} JO a1njeIs 94} ‘OOZ-OGT ‘saeq -WNnU [e}UOZIIOY AAVOY JO MOI JOMOT 9G} PUL S20ISOU o[BUI JO JequUINU 9yN[OSGe oY} Sjusseider ‘OZ—-O ‘sdoquinu AAvEY [[VUIS JO UUIN[OD [COT}IOA OL *s]0103] 0} JO 91njv4s ou} ‘SLI-FEl ‘Sdequinu Jo Mod [vjUOZTAOY, OSV] OY} PUB ‘SJOIOS] JO AaquINM sjN[osqv oY} syUoSeIder ‘FT—-O SdequInuU JO UN[OO [vdOTPAOA OBIe] OYL—T “Yl Go our 891 1 -t 0 422 BEAN. ing of types. ‘This meaning is obscured by constructing a curve with 1,000 individuals from the same population. Take for instance the height of about 150 negroes that I measured at the Johns Hopkins Hospital Dispensary in 1906. There is great irregularity in the curve, which has seven summits, as may be seen in the small figure in the chart. Each of the summits represents a group of negroes well recognized in the United States, and known to exist in about the proportion and with the height given. Pygmies are rare, Hottentots are not plentiful, and extremely tall negroes are seldom found. The Guinea Coast negro, the Kaffir and the mulatto make up the bulk — of the negro population in America. With 1,000 or more individuals, the types as represented by the height would be obscured and only an average negro height would be the result. Topinard(62) classifies 48,282 negroes and mulattoes by height and arrives at a single result, to wit: The average of the American is the same as that of the African negroes, namely, 168.1 centimeters. This is a not insignificiant result in itself, but it leaves much to be desired in the classifi- cation of negro types. Much depends upon the selection of individuals, but in a random sample with no selection there should be no spurious types. There was no conscious selection of negroes or of Igorots, but every available individual was measured. Three summits are evident in the curve of height for the female Igorots as well as for the males. The summit for tall women (160 centimeters) is considerably prolonged, while that for small ones is not. The inter- pretation of the curve would be as follows: The large central portion represents the majority of the people, and this is the most frequent type. The two extremes of the curve represent a small and a tall people who have mixed with the others. The small people are few in number and have had slight influence in altering the type, while the tall ones are in greater numbers and have modified it considerably, causing the central part of the curve to be flat-topped by increasing the number of individuals with height above the mean. Reasoning from this premise we may conclude that the original stocks of Igorots had a mean height of about 150 centimeters or less, which is the same as that of the people of the inland part of the Malay Peninsula(26). Martin’s curve, however, shows three summits for both males and females, and there is evidence of three peoples among his subjects, so one must search back of this for the primitive stock. Whatever that may have been, the influence of a tall people is evident, and this came at a remote time, when the tall people were present in great numbers. Later came the influence of the small people, which there is good reason to believe were the Negritos. The mean height of 10 male Negritos of the Philippines according to A. B. Meyers is 144.5 centimeters, with extremes of 140.1 and 150.5 centimeters respectively (29) ; Montano gives the mean height of 18 male Philippine Negritos as 148.5 centi- meters; Deniker(14) presents 42 Aeta-Negritos with a mean height of 146.5; Keane(21) gives the mean height 147.3; and Reed(47) states that of 48 mixed Negritos to be 146.3 centimeters. The height of 4 adult male Igorots is 142 centimeters, and there are 32 below 150.5 centimeters. The height of 5 male adults is nearly 170, and 60 per cent are above 150 centimeters. It is easy to conceive that a few Negritos would become attached to the Igorots in their progressive conquest of the mountains, but it is not so easy to believe that a tall people has joined them in the Philippines to make up about one-third of their number. Stray refugees or adventurers may have come to them from time to time THE BENGUET IGOROTS. 423 in their mountain fastnesses, as in the case of an Ilocano, 168 centimeters in height, who came to Atoc from Union Province at the age of 15 years, fought in the war parties of the chiefs and was accepted as one of them for his continual daring and bravery. He is taller than the average Igorot, but not above the height of three of those measured. The Spanish influence must be reckoned with because the Spaniards have been in contact with the Igorots for at least fifty years (7), although no individual measured showed any indication of the Spanish influence in physical characters. A few tall men added to the Igorots from time to time may have had a slight influence, but they could not have altered the average height materially; furthermore the inaccessible parts have the tallest individuals, - and the Bontoe Igorots; the most inaccessible and remote, are the tallest of the Igorots. II. PROPORTION OF THE BODY PARTS. The measurements made on the living are necessarily more inaccurate than those upon the skeleton, but with proper precautions and great care they may be used as differential factors in the physical anthropology of a people. UPPER EXTREMITY. (Eatremitas superior.) The Igorots are essentially short-armed, although there are long-armed individuals and the several groups show differences in the absolute length as well as the relative length. The mean (absolute) length of the upper extremity of 104 adult males is 67.82 centimeters, which is less than that of any other related Malay peoples, except the Senoi group(30) in the Malay Peninsula, and it is little less than that of the Japanese. The length decreases progressively with locality and altitude from the high- lands to the lowlands as may be illustrated by grouping to show the mode and the extremes: Absolute length of upper extremity, in centimeters. f 7 5 Group, sex and age. | 40-49. | 50-54. | 55-59. 70-74. | 7). Total. | | | 60-64. | 65-69. | ; — a | Wowie (besstetal Pssas |Saeeaes eae) Bf 8 a | Highlands___________ LE Ce Se Ee eee | 5 26 | 4159 46 Lowlands Op 1D 22 | fig ee ome 44 | Adult males_______ 2 | 17 51 | 31 | 2} 104 Woomlensie: 2 20k) So ote, 2] 5 | 3 | eee eens | 10 Boys, 12 to 15 ------_-|___-:_-- 1 7| - 10 lg |ecesnese BSc ser 19 | Boys, 10 to 12-__-____| 1 4 2 | | 7 | Boys, 10 and less____. | 3 Dili sxe ses ———— | Seog ed pe SUE | } 1 The range of variation judged by the difference between the extremes, and the spread of the mode, is greater in the lowland than in the highland or Bontoe groups, which indicates greater diversity of type in the lowlands. Reasoning from this premise, the conclusion is that Igorots from the lowlands are more recently mixed than the people from the highlands or from Bontoe. The variation of the upper extremity as expressed by the difference between 424 BEAN. minimum and maximum, is 31 per cent of the maximum length for the entire extremity, 21 per cent for the upper arm, 30 per cent for the forearm, and 40 per cent for the hand. The actual variation in centimeters is: Centimeters. Entire extremity 23.1 Upper arm 7.4 Forearm Weil Hand 8.0 The upper arm which is least variable is therefore a better factor for testing type differences, and the greatest difference is found between the Lowland and Highland Igorots in this part. Its mean length for the Bontoc Igorots is 0.8 centimeter greater than that of the highland group, and 2.0 centimeters greater than that of the Igorots of the low- lands. The mode is 1 centimeter greater for the Bontoc Igorots than for the highland people and 4 centimeters greater than for those of the lowlands. (Tables V, VI, and VII.) ‘ According to this standard and by the total length of the upper ex- tremity, the Lowland Igorots correspond to the Senoi of Martin(31), the Highland and Bontoc Igorots to the remainder of the population of the Malay Peninsula, as this table indicates: Mean lengths of upper extremity, in centimeters. Entire Armminus Upper + extremity. | hand. arm. | Forearm. Hand. Group. = IF = as Abso-| Rela-| Abso-| Rela-, Abso-| Rela-| Abso- Rela-/ Abso- Rela- lute. | tive. | lute. | tive. | lute. | tive. | lute. | tive. | lute.| tive. | | | | | | ee ane: Senolees eens 66.7 | 43.1 | 50.5 | 32.3 | 28.0 | 18.1 | 21.3 | 13.8 | 16.8} 10.9 Blandas ______--- 69.9 | 45.4 | 52.5 | 34.0 | 30.3 | 19.5 | 22.2 | 14.3 | 17.4} 11.3 Malayen_____---- 71.8 | 45.9 | 53.1 | 34.0 | 30.4 19.1 | 22525) 13.8 (1852) |) de Lowland —____-_- | 66.5 | 43.8 | 50.5 | 33.8 | 28.6 | 18.7 | 22.0] 14.5 | 16.2] 10.6 Highland ______- | 68.3 | 43.9 | 52.1 | 33.6 | 29.8 } 19.8 | 22.4 | 14.4 | 16.0} 10.3 IB ONTOC=== a= 71.8 | 45.2 | 53.9 | 33.9 | 30.6 | 19.2 | 23.3 | 14.8 117.8} 11.3 | Tgorot women ___| 62.8 | 42.7 | 47.6 | 32.4 | 27.1 | 18.5 | 20.5 | 14.0 | 15.1 | 10.3 | I | | | | ! Martin’s Senoi group is almost exactly the same as the Lowland Igorots in the absolute and relative length of the entire upper extremity, the upper arm and the hand, and in the absolute length of the arm minus hand, and the forearm. The Igorots from the lowlands have a relatively longer forearm, and a relatively longer arm minus hand than the Senoi. Martin’s groups of Blandas and Malayen are similar to the Highland and the Bontoc Igorots in every measurement, except that of the forearm (absolute and relative), wherein the Bontoe Igorots exceed all others. The relative forearm length of the Bontoc is 14.8 centimeters which is equal to that of certain Europeans, below that of other Europeans, and considerably less than that of the negro(34). It corresponds to the Me- nangkabau-Malayen and the southern Chinese of Hagen(33), and oc- THE BENGUET IGOROTS. 425 cupies a position midway between that of the Japanese and the South American Indian. e The relative upper arm length of the Highland Igorots, which is the same as that of the European (19.8 centimeters) is as great as that of any other people so far measured except the Sikh (20.1 centimeters) (33). The relative hand length of the Bontoe is also the same as that of the European, and the relative length of the entire upper extremity is but a trifle less. That of the parts of the upper extremity and of the entire upper extremity place the Bontoe Igorots nearer the European than are the Highland or Lowland Igorots, and the Highland are nearer than the Lowland. The hand of the Highland, and of the Lowland as well, is unlike the European, Chinese, or negro, because it is relatively shorter. It is the same as that of the Igorot women, but less than that of any other people except the Senoi of Martin. The absolute and relative length of the entire upper extremity, and of each of its parts is slightly less for the women than for any group of men. The hands of the women and the men of the highland group, however, are exactly the same in relative length. The ratio of the forearm to the upper arm, the so-called “brachial index,’ is important to establish the relationship of the Igorots to other people. This index is 76.2 for the adult male and 75.6 for the adult female Igorots. It is 76.9 for the Lowland, 75.2 for the Highland, and 75.1 for the Bontoec. The brachial index is an additional differentiating factor for the Lowland Igorots, placing them in the same class as the Senoi, while the highland and Bontoe groups in this factor, as in so many others, are more like the European. The brachial index of the living has been determined by many different methods, and by so many different authors, that divergent results are reached on the same people. For instance Sarasin found the index of the Veddahs to be 91.9, while Martin determined it to be 73.8(32); Weisbach gives for Germans an index of 83.5 and Teumin for the Russian Jews one of 72. As Martin remarks there is great need for a fundamental reform in the methods of measuring the living. The brachial index on the skeleton is greater than on the living, and is given by Martin for Europeans as 72.5, for Negritos, 83, and for the Senoi, 78.9. On the living, Martin gives the Senoi, 76.0 and the Blandas, 73.2, which again places the Senoi and the Lowland in the same class, while the Bontocs are nearer the Blandas. I LOWER EXTREMITY. (Extrenvitas inferior.) The measurements of the parts of the lower extremity are more exact than those of the upper, because in the latter there may be unconscious and unnoticed shifting of the parts when the measurements are being made, whereas in the former there is greater stability because the parts are placed firmly on the ground. The height of the pubis taken from the pubic spine is more accurate than the height of the trochanter, because of the accessibility and the ease with which the spine is located, whereas the heavy fascia, muscles and ligaments over the trochanter interfere with exact work. Parallel measurements for comparison indicate that the pubis is 426 BEAN. slightly lower than the trochanter in the Igorots, but this may be due to the heavy gluteal muscles and fascia lata occurring in these incessant mountain climbers. Height of trochanter and pubic spine compared, in centimeters. | Group. Number. | Trochanter. | Pubic spine. | [Een =| | | Bontoe ~~ ------------ -----an2-a222------- | 14 82.0 81.9 | | Highland 46 79.9 77.9 | | Lowland 44 78.1 76.0 | Mota oases ean lat Le aig Ree ae 104 79.4 77.6 | | WOUND GIL 5 a a a a ae 10 74.1 71.7 |’ Here again as in the other measurements, the Bontoc is the greatest, the Lowland least in absolute length, and the highland group is between the two. These measurements represent the absolute length of the entire lower extremity, and a glance at the next table will show that the rel- ative length follows slightly the absolute length. Again there is the similarity between the three groups of the Malay Peninsula(35) and the Igorots. Length of lower extremity, in centimeters. | at Approximate height N JUAN, of ankle. Group. Lees Absolute. | Relative. | Absolute. | Relative. i} lay BOntOC === | 14 82.0 51.7 5.6 3.70 Mighlandee se eae e ae 46 79.9 51.6 5. 2 3.55 |peuowland)=s2ncee-sanecteaee 44 78.1 51.5 Gal 3.50 Totalstes bee asses 104 79.4 51.6 5.4 3.55 Womens 3 10 74.1 503533 | Pees Saeko | Eee ees The lowland group again corresponds to the Senoi, the relative length of the Bontoe lower extremity is almost identical with that of the Euro- pean, and is less than that of the negro, but considerably more than the Japanese. The women have absolutely and relatively shorter legs than the men of the highlands, with whom they should always be compared, because they belong to that group. The length of the leg-minus-foot is approximate because the ankle height was obtained on but thirteen Igorots. However, the average of even so small a number emphasizes a fact that I observed constantly, to wit: The distance from the internal condyle to the sole of the foot is so short that it could with difficulty be measured, especially where the surface of the ground on which the individuals stood was rugged. The ankle height is similar to the hand length, especially in so far as the shortness is more pronounced in the highland and lowland group than in the Bontoc. THE BENGUET IGOROTS. 427 Length of leg-minus-foot, in centimeters. = | ‘i Group. Number. |} Absolute. Relative. | | | | | EBON TOC SaaS ee Se aE So ee ee 14 | 76.4 48.0 | Highland____ 46 | 74.7 48.1 | Lowland ------ 44 73.0 48.0 | ETT tal Leet ed eed est aban GI ST ae Se | 104 | 74 | 48.1 | L | i | The absolute length of the leg-minus-foot follows the stature closely, but the relative does not, since it is equal in all the groups. The intermembral index may be calculated from this, comparing it with the arm-minus-hand, but a fairer consideration under the cireum- stances would be a comparison of the entire lower with the entire upper extremity. Both are given so that the intermembral-index may be com- pared with that for the inhabitants of the Malay Peninsula(36). Intermembral index. ] | Arm-minus-bhand-vs. | Entire extremity, | é leg-minus-foot. | upper vs. lower. Group | Ne | 2 | ber. Bean. Martin. Bean. Martin. Bontoc= esses 14| - 70.6 | Blandas 71.4 | B. 87.5 | Blandas 87.3 | Highland__ | 46 70.0 | Besisi 70.1 | H. 85.4 | Malayen 85.1 Lowland -_ 44 69.2 | Senoil. 69.7} L. 85.1.) Senoil. 84.1 | Potala 104 OO), | eae ae BO 54g) Boe eee WONG Na ST () | a ee ee ee 84.7 Eee eee The Igorots have relatively longer arms as a whole in proportion to their legs than the Malays of the peninsula, but there is a general con- cordance as usual, and a similarity exists especially between the lowlands and the-Senoi. The arm of the women although 5 centimeters absolutely shorter than that of the men, is relatively to the leg, about as long. For the parts of the lower extremity the absolute length of the upper leg (femur) is the difference between the height of the knee and the trochanter, and the absolute length of the lower leg (crus) is the dif- ference between the height of the knee and ankle. Length of the parts of the lower extremity, in centimeters. Upper leg. Lower leg. | Tibio- 3 Num- : Group. Phare |= feng Stature. Absolute.| Relative. Absolute. Relative. index. Bontoe _-___-_- | 14 39,4 24,9 36.9 23.3 | 93.6 | 158.6 Highland__ -| 46 38.3 24.8 36.3 23.4} 94,7 | 154.9 | Lowland _____- | 44 37,2 24.5 35.8 23.6 | 96.2 | 154.6 Women____-_- (10 33.9 PEA [eee el eerste | Sesame 146.7 | | Total ____| 104 38.0 24.7 36.0 23.4 94.7 | 154.0 | 428 BEAN. All the lengths show in general the same relation between the groups of the Malay Peninsula and those of the Igorots as the preceding measurements, except that the tibio-femoral index is from five to ten points higher for the Igorots. This corresponds to the crural index of Hagen(36) for the Malay and Melanesian people, which is 90.8 “bei Alas,” and 97.4 “bei Neu-Mecklenburgern.” The measurements are equal to those of the European and Japanese, and are less than all other related East Asiatic people except the Aino. The length of the upper leg, both absolute and relative, is less for the Tgorot women than: for any of the women from the Malay peninsula. The relative length of the upper leg follows the absolute, as does the relative length of the entire lower extremity, so one may say the length of the lower extremity is determined by the length of the upper leg, which in its turn determines the stature of the individual. In other words the correlation of stature and length of upper leg is pronounced. The same is not true of the lower leg, but rather the opposite. With absolute increase of length of the lower leg, the relative length decreases, so that the shortest Igorots have relatively the longest lower legs. The tibio-femoral index presents this clearly. Compare it with the brachial index, and a striking similarity between forearm and lower leg is noticed. The shortest individuals have relatively the longest forearms and lower legs, while the longest individuals present the reverse. It may be of interest in connection with the body parts to present the absolute dimensions of an Igorot and of a Senoi man(38). Select- ing an individual Igorot with the same height as the Senoi man, the body parts correspond almost exactly, except that the hand of the Igorot is shorter and the upper arm longer than the same members of the Senoi. This is corroborative evidence in a special case of the general evidence obtained from the averages of the body parts. (Fig. 2.) 153 152.6 220 joo 66.1 67.0 26 2g 16.8 6 16.9 403 40.2 80.1 806 3a) 35.2 Fic. 2.—Absolute length (in centimeters) of the body parts of a Senoi man (left figure) and an Igorot man (right figure). THE BENGUET IGOROTS. 429 The absolute and relative length of the extremities of the Igorots may be stated in contrast with similar measurements of other people. The absolute length of the upper extremity of the Igorots is less than that of any other Eastern Asiatic people except the Senoi. The Northern Chinese have the greatest and the Japanese an intermediate length, or nearer the length of the Igorot arm. ; The relative length of the upper extremity of the Igorots is less than that of the majority of the associated people. However, the Japanese and the Senoi have this relation shorter; the former the shortest of all, and the Veddahs and Ainos the longest, with the Europeans intermediate. The Igorots are midway between the Europeans and the Japanese. The absolute length of the lower extremity of the Igorots is less than that of any other except the Japanese and Cochin Chinese, who have this dimension slightly less than the Igorots. The Japanese have the shortest and the negroes the longest legs of all people, while the Europeans are intermediate. The Aino and the Igorot are almost identical in leg length, and are midway between the Japanese and the European. Martin’s observations lead him to conclude that relative leg length and absolute leg length follow each other closely, in which I agree with him. The relative leg length of the Igorot is slightly nearer the European standard, but otherwise corresponds to the absolute leg length. SHOULDER—HIP-PELVIS. The shoulder width (acromian), and the width of the hips (iliac) may be contrasted and compared with the height, to determine relation- ships and differences. Relation of the (iliac) width of the hips to the (acromian) width of the shoulders. Sse plete eae ee see et vee Shoulder. Hip (coxa). | Relative Pelvis. Num- hip peer 10s Group. peed ber. Abso- | Rela- | Abso- | Rela- shoulder Ant.- | Pelvic lute. | tive. lute. | tive. | breadth.) post. | index. Bontoe 14 34.9 22.0 26.4 16.6 75.6 Highland _- 46 35.0 22.6 26.4 17.0 75.4 Lowland 44 34.4 22.7 25.1 16.6 72.9 | Kren chiment = sss=s=ssa—eaee French women_-_-__----------- Belgian men ____- Se as Belgian women _-_---__---_-- Negro men European men —_--_---_----._ * Topinard (66). to 430 BEAN. The shoulders of the men are relatively (to the stature of course) wider than those of the women and the Lowland men have the widest shoulders. The hips of hip the women are relatively wider than those of the men. The breadth is shoulder naturally greater among the women, but this is much less for Igorots than for Europeans. The Bontoe and Highland Igorots approach the European more closely than do the Lowland. The difference between the Bontoc and the Frenchmen is 5.2, the difference between the Lowland and the Frenchmen is 7.9, and the difference between the Igorot women and the French women is 10.6. The reason for this disparity on the part of the Igorot women is not so much in poorly developed hips as in well-developed shoulders due to field work and burden bearing. The Igorots are intermediate between the European and the negro in relative hip breadth. j The pelvic index is given for the lowland group alone, because it was determined for no other. THE UMBILICUS. The position of the umbilicus in relation to the pubis and the supra- sternal notch, although it is more variable than such fixed points as the two last mentioned, is of importance in type differentiation. Its im- portance embryologically can not be denied, but whether its position is due to developmental phenomena or not, remains to be determined. I present for the first time the mdex of the umbilicus, and emphasize its singificance. The index is found by dividing the distance of the umbilicus from the pubic spine by its distance from the supra-sternal notch. This in- dicates its relative position on the body. If the index is high, the umbilicus is relatively near the suprasternal notch, but if low, it is relatively near the pubic spine. I propose the name of omphalic index for the index of the umbilicus. Divisions into hyper-, meso-, and hypo-omphalic would follow naturally for the high, intermediate, and the low umbilicus. It is mexpedient at this time to attempt a definition of the limits of these three classes of omphalites, although I am inclined to believe that the Igorots are hypo- omphalie. OMPHALIC INDEX. | | isan | Sternum| Pubis to | Index of | Group. Pubis. 05 Sternum.| to umbi- umbili- | umbili- (ec s all | liens. cus. | cus. | v i | | Bontoe —___..-__ ee si)! 95.6 129.6 34.0 13.7 40.3 Highland _ 91.9 126.5 34.6 14.0 40.4 Lowland ___ 89.8 122.7 32.9 13.8 41.9 Total 91.5 125.3 33.8 13.9 41.1 Highland women ___ 71.7 87.6 119.4 31.8 15.9 50.0 i { The sexual differentiation by the omphalic index is great. The women have an index that is 25 per cent higher than the men. The differences THE BENGUET IGOROTS. 431 between the three groups of men is not so marked, but the lowland group resembles the women more than any other. The tall men have a low and the short men have a high index. Childbearing in women may have some influence on the position of the umbilicus. So may the protrusion of the abdomen from any cause, such as obesity, ascites, rice feeding, etc. The relation of omphalic index to age is as follows: Omphahe index and age. Imbili | Sternum Pubis to | Group-age. wun | Pubis. Gant Sternum to umbi: | obits Preset Below 10 ___- 5 54.3| 64.3 | ge7| 22:4] 10:0 4406 10S Se 7 61.1 71.3 | 98. 2 26.9 | 10.2 38.0 AQ=13 Soca 6 68.2 79.5 107.6 28.1 11.3 40,2 14=155e225 os 13 70.5 | 83.0 | 114.0 31.0 | 12.5 40.3 | 1 9 76.8 90.8 | 124, 2 33.4 | 14.0 41.9 | | The index is high before the age of ten and decreases thereafter. The decided drop at 10-11 may be erroneous. At the age of 16 the position has reached that of the adult. The position of the umbilicus in the small male child is similar to that of the woman. BODY LENGTH AND NEOK LENGTH. The stature may be divided into four parts: Head length (chin to vertex), neck length (chin to sternum), body length (sternum to pubic spine), and leg length (trochanter to sole). The leg length has been given, the other three remain. The body length is said to be 4 centi- meters less than the distance from the suprasternal notch to the pu- bis (37), and the upper end of the leg is parallel with the lower end of the body, or 4 centimeters above the pubis. However, I find only 2 centimeters’ difference between the pubis and the trochanter and as the pubic spine is more definite than the trochanter I prefer to use the spine. The body length from the supra-sternal notch to the pubic ‘spine is as follows: Body length (truncus). Group. Absolute. | Relative. Highlan Gee o 2 ose saa ee eee econ neee eae 48.6 31.3 Rowland ee eee oe ene ee ee eo 46.7 30.8 | | Bon toc). 2 Sees ahs ee ee ee eee Se 47.7 30.0 | WOM ens nee ae ae Soe en ae ee eed nee eee Et 47.7 32.3 Faye ig eee eres eer sees SUA HE oer Me ae eC 47.7 31.0 | 432 BEAN. Compared with Martin’s figures for the Malay Peninsula, the body length is slightly greater for the Igorots, and especially is this true of the highland group. The lowland is exactly the same as Senoi II, in absolute length, but relative to stature, the Lowland Igorot has a longer body. The body of the women is exactly as long as that of the men, and relative to stature it is longer. The neck length presents unusual differences. The neck of the High- land Igorots is the shortest, even shorter than the women’s, both abso- lutely and relatively; that of the Lowland Igorot is longer than any others, even the Bontoc being shorter. Neck length (collum). | | Relative. | Group. Absolute. The women have relatively as long necks as any of the men; although they are not exactly swan-like, there is symmetry and beauty in their lines and proportions. Ill. HEAD FORM. The length of the head is measured from the glabella to the maximum occipital point (torus occipitalis), the greatest breadth is taken, and also the height from the external auditory meatus to the bregma. The following outlines are made with electric fuse wire: 1. Sagittal: From the glabella to the external occipital protuberance; 2. Horizontal: Above the superciliary ridges and around the maximum occipital point ; 3. Coronal: From the root of the zygomatic process on each side across the vertex. Electric fuse wire was chosen after trying many materials, because of its lightness, rigidity and pliability. If care is exercised, the shape of the head is retained perfectly, the hair interferring in only a few in- stances. This method of obtaining outlines of the head on the living, while open to objection, nevertheless furnishes a ready and convenient means of securing at least the approximate head form. The sagittal outline is especially valuable, because of its greater accuracy. The measurements of the head are reduced to skull measurements by deducting 10 millimeters from each diameter, and the skull size as thus determined is used THE BENGUET IGOROTS. 435 in the calculation of the cephalic index. The breadth ji ndex is then classified as length follows: Hyper-dolichocephalic 70 and less. Dolichocephalic 70 to 74.9 Mesocephalic 75 to 79.9 Brachycephalic 80 to 84.9 Hyper-brachycephalice 85 and more. Aurel von Térék’s(68) classification is useful in determining the actual size of the skull and its length in connection with the Taste index. He uses symbols which are the initial letters of the three groups representing the three sizes of length and the three of width of the skull, namely, small, medium, and large. These will be symbolized by their English equivalents, s, m, and 1 as follows: _———— — —E__ = ee, _ a — | _ s=narrow eee aS | l=wide (short), |™—=me dium. Gong) | | mm. mm. | mm. | Greatest skull width varies 101 to 173 mm____-_________ | 101-125 126-149 150-173 | Greatest skull lenght varies 143 to 224 mm _____________ | 143-169 170-196 197-224 | | Each skull is given a number, 1 to 82, which corresponds to its length in millimeters; No. 1=143 millimeters and No. 82=224 millimeters. Cephalic index—male adult Igorots. Type. m/s m/m 1/1 | Total. = | e ee 2S a ail isle ee Se alien | % if | Number -_------ 12 14 16 18 20 22 24 26 | 28 30 32 34 36 38 40 42 44 46 48 50 52 54 | 56 58 | | Dolichocephalic 1 i ho" 8 Oh 2 8 [eee et |} Leh | Mesocephalic___, 1}).8P 669 268.9 2-7 [ieee | 43 Brachycephalic | 1 I Seats PAA) eal Baa ca DNS SAS Te TOR Bea se ia eae | 18 | 1 hyperbrachy- | cephalic at | | | | 30; 1 hyperdo- | | | | lichocephalic | | | Qt a nc | ce Re Oe Dea | 2 | Let the symbol above the line represent the head width and the symbol below represent the head length, then the classes fall almost entirely in the m/m group or medium-sized heads. === —S Class. [Rum sis | m/s | s/m | syn Ym | mj | a | ‘| | | = = Seal | CHEM ER aa Ea —-|——— | Hypérdolichocephalie __- yo @ | 0 My} 2 | 0} 1 @ 0 | Dolichocephalic___----____-___ | 41 iy] 0} 0, 40 0 0; Oo Mesocephalic_________---______ Wenee 43) |TeetOU ere 0 42 fh Oh Or @ Brachycephalic _-_----__-_-___ } 1B OU ® | ay o| o { i} 1 i | Hyperbrachycephalic —_______- | 1 0 | 0 | 0 | 1} 0 | 0 0) | otal cme cen i I my al el oO} ey Of Oo | | | 434 BEAN. There are no heads larger than medium size, but there are 9 smaller ones, 7 of which are brachycephalic. The smallest head of the series is dolichocephalic. There is a preponderance of dolichocephalic heads [41] over brachycephalic [18] which indicates that the Igorots are largely a dolichocephalic people, with medium- sized heads and that the brachycephalic portion of the population has small heads. A more detailed analysis reveals the relationship of the head form in different localities. It is to be seen that the Bontoc group is funda- mentally dolichocephalic, the highland is largely dolichocephalic and mesocephalic, while the lowland is for the greater part mesocephalic and brachycephalec. The percentage of dolichocephalic heads decreases from 57 in the Bontoc group to 29 in the lowland; that of mesocephalic heads increases from 29 in the Bontoc to 46 in the lowland; while the per- centage of brachycephalic heads increases from 14 in the Bontoc to 25 in the lowland group. It is of interest to note that 54 per cent of the brachycephalic heads in the lowland group belong to the m/s class, 46 belong to the m/m, and 86 per cent of the m/s brachycephals are in the lowland group, whereas only 14 per cent are found in the highland and none in the Bontoc. The Bontoe group has not only a larger percentage ot dolichocephalic heads and a smaller of mesocephalic and brachycephalic than the other groups, but there are no small heads in the Bontoe group. Cephalic index by locality. { = r = ! } . 5 | | Num- Per Group. Class. | ner! s/s m/s | m/m Rent Bontoe -___---- | Dolichocephalic___------__- pe Bibs Os: Ble | Mesocephalie_____-_-----___ | 4 | 0 0! 4] 29 Brachycephalic ____--_--__- | 2} 0 0 2 14 | : } ——- ——- —_— Motel ia Sites te re a 0 0 14| 100 Highland ____- Hyperdolichocephalic _- Thar wo 2 | Dolichocephalic_____ 20 | 1] 0} 19 44 | | Mesocephalic______________- leu 93 0 0 19 41 | | Brachycephalic --_-.-------, 5 | 0 1 4 li Hypherbrachycephalic ae) 1 0 | 0 1 2 | Tic talento | 46 | | | ico) Lowland _____- Dolichocephalic__________-- 0. 0. 13 29 | Mesocephalic_____- 0 1 19 46 | . | Brachycephalic _— 0 6 5 25 | | Tote sae ane 0 7| 37| 100 | | It has already been demonstrated from von Toérdék’s classification of the cephalic index that neither the length nor the breadth of the Igorot head is above medium size, and at least one group, the Lowland, has head dimensions below medium size. THE BENGUET IGOROTS. 435 Head dimensions of the living. | | N Tele Breadth, | Auricular| Height Group. hare length |bregmatic) length | \Length.|Breadth.| imdex.* | height. | index. BONO Cheese anaes 14 18,8 14.5 V7.1 Te} it} 70.0 Highland =sssss seen 46 18.9 14.7 77.8 | 13.3 70.4 owl ancien saan 44 18.6 14.6 78.5 | 12.7 | 68. 3 | Potalia=2- sss ss 104 18.8 14.6 77.6 2.9 68.6 WOM GM = ss eae 10 18.2 14.1 77.5 | 12.3 67.6 | Bontoe (Jenks, 19) -__- 32 19.2 1552) TOT | reeeee sone bese eeese 1 ! = ! “The index should be reduced from 0.5 to 2 points to equal the skull index. The Igorots are more dolichocephalic than the eastern Asiatic people, breadth length index is dolichocephalic and slightly mesocephalic and the Lowland is breadth length height index is similar to the breadtls in its relationships, but length length the women are less high headed than the men. The Bontoe and high- land groups are higher headed than any other people of the Malay Peninsula or of eastern Asia, except the southern Perak Malays(1) who are four points higher. As the Lowland Igorots, who are largely brachycephalic, have low heads and the Bontoe Igorots who are dolichocephalic have high heads, it is to be presumed that dolichocephalic, or long heads, are high, and brachycephalic, or short heads, are low. However, the reverse is known to be true, although when I first noticed the group variations, I thought the Igorots might be different in this respect from other people, but but less so than the tribes of India. The Bontoe and Highland mesocephalic. The women have the same index as the men and the : breadth . 3 height : comparing the -——— index of the head with the ——"— the result is length length as follows: : breadth ._ height . Comparison of with g mdex of the head. length length | Tndexoh los =e dee ae 63 64 65 66 67 68 69 70 71 72 73 74 75 | Dolichocephalichesss= === ae WB By Ge Bh wy 1 1 | | Mfesoceph ali cies ee eia man emul nanne Lili@466788 8 1] | Brachy.cephalicr = ssa === | ab ab al 2h Db) By | 436 BDAN. The dolichocephalic heads are low, the brachycephalic are high and height length of the lowland group be accounted for when it is known that this group is largely brachycephalic? A closer examination of the cephalic index reveals the fact that the dolichocephalic heads of the lowland group height length so as to lower it. It is seen from the above table of comparative indices that the a index of the dolichocephalic heads is grouped about 65 5 and 69. There are high long heads and low long heads, the latter are found largely in the lowlands, and the former are found largely in the highlands. The widest head breadth compared with the narrowest forehead breadth gives a great difference between the Bontoc and lowland groups. The Bontoc, with the narrowest head, has the widest forehead, and the Lowland with a wider head, has the narrowest. The women have rel- atively wider foreheads than the men. The forehead of the Igorots is wider than that of the inhabitants of the Malay Peninsula, or of other Malays, as wide as the Northern Chinese, and a little less wide than the Aino (39), index the mesocephalic are intermediate. How, then, can the low have a very low index which influences the average of the group Widest head breadth compared with narrowest forehead. | Widest | Narrowest | Group. | part of partof | Difference.) head. | forehead. | ).: (BONtOG. S34 te eee Ie ee ee Se Sealing! S 14.5 | 10.5 | 40.0 Highland’ 2202 sae ee a 14.7 10.3 | 44.0 Lowland:222¢ e225 ee eee i 14.6 10,25 | 43.5 | | Total eee Fee eee Pease era ee teen | 14.6) 10.3 | 43.0 | “AWiom en = <2! cesses ee ee eee 14.1 10.3 38.0 IV. PHYSIOGNOMY. The morphologic face height is the distance from the chin to the nasion; and the physiognomic face height is from the chin to the hair line. The dimensions are practically the-same for the Bontoc and high- land groups, but the lowland group is smaller in every particular. The face width of the Bontoc and Highland is greater than that of the in- habitants of the Malay Peninsula, and is nearer that of the Chinese, Japanese, and Ainos. The physiognomy of the women is less in its dimensions than that of the men. ~ THE BENGUET IGOROTS. 437 Dimensions of the face. | Phy foe! oe pia : Num- “homie “Morpho | Bizy-_ | Physiog- ) Morpho- { Group. ise | gage Pee gomatic nomic | logic | Es height. height. | width. | index. index. | : eed Eni |e MOEA Ti NOI BOntOC Hae =eennen 14 | 18.1 10.8 | 13.7 75.7 78.8 Highland __--_______ | 46 | 18.1 | 10.9 13.8 76.3 79.0 Dowland | 44 17.5) 10.7) 13.8 76.0| 80.4 | 104 | 17.9 10.8 | 13.6 76.0 | 79.4 10 16.6 10.3 | 13.1 79.0) 78.6 | | i The index of the physiognomy, which indicates the relative face width, is greatest for the mountain division and least for the Bontoc. It is greater for the Igorots than for the Japanese and Malays, but it is less than that of the Aino. The women’s faces are relatively wider than the men’s. The morphologic index which indicates the relative length of the face below the eyes is greatest for the Lowland and least for the Bontoe. It is less than that of any other Eastern Asiatic peoples, although the Mantra(49) are about the same. The lower face height (chin to nasal septum) as compared with the artistic modulus and with the total head height (chin to vertex) is as follows: Lower face height compared with the artistic modulus and total head height. | Relative Total Artistic | LOWer | lower face Group. | Num- | head face height to | per. | height. MOUS.) peieht. | total head | | height. | | Bontoc | 14 21.8 | 7.3 | 6.7 | Syl | Highland_ 46 22.0 7.0 6.6 30 | Lowland ___ | 44 | 21.1 oat || 6.7 32 | Total - ee |) 216 al 67 31 Women 10 2057 7.1 6.5 31 | | The Highland has the least lower face, the Lowland the greatest, and the Bontoe and the women are exactly intermediate. The Highland has the greatest total head height, the longest physiognomy, and the shortest lower face, therefore his frontal cranial height is the greatest of all the Igorots. This is true also of the auricular bregmatic height, and the head outlines show the same, therefore the several measurements corroborate each other. The artistic modulus of the Bontoe is nearer that of the Europeans than are the others. ‘ 438 BEAN. NOSE. The nasal dimensions considered with those of the mouth are given in the following table: Nasal and oral measurements on the living. Num- | Nose, | Nostril,| Nasal | Num- | Mouth,| Lip, Group. ber. |length.|breadth.| index.} ber. |length!| width. | Mouth, length.» | |) Bontoc)-=a es | 14 | 4.1 4.0 97.6 6 4.4 1.1 | Chinese 4.7 Highland __________ 46 4.3 | 3.8] 88.4 35 4.8! 1.2) Parisian g, 5.0 Lowland ___-_-_____ d4 4.0 | 3.8 95.0 28 4.9 1.2 | Parisian 9, 4.7 Motaleesseen 104) 4.1 3.8 | 92.7 69| 4.8] 1.2| Negra 5.31 | Women!2=222e: 10 3.8 3.8 | 100.0 10 4.4 | 1 Negress 5.1 Bontoe (Jenks 19) __ 32 5.3 AYO 3782) eer een | cee eae Sa eee eres LSS aTopinard (63). The measurements of the mouth may be dismissed by stating that the lips of the Igorots are full, but not thick and protruding like those of the Negro, nor is the mouth so large. The Bontoc and the women have smaller mouths than the Highland and Lowland Igorots. The height of the nose measured from the subnasal point to the nasion is 7 milli- meters less than the average height of this feature of the inhabitants of the Malay Peninsula(41), and the breadth (ala nasi) is the same, whereas the height and breadth are but 2 millimeters less than that given by Annandale and Robinson(48) for the same people. The resulting nasal index is therefore 10 per cent greater than Martin’s for the Malay Peninsula and about the same as that of Annandale and Robinson(1). The extremes of nasal index found are 72 and 115.* The women of the Malay Peninsula haye narrower noses than the men, while the Igorot women have wider: It may be of interest to note that the nasal index of the dolichocephalic Lowland Igorots is 99.4, while that of the brachycephalic is 85. This would seem to indicate that there are two types of Igorots in the lowlands, the long headed beimg wide nosed, the broad headed not to such a degree. EYES. The eyes are measured by taking the distance between the inner corners (commissura palpebrarum medialis) and between the outer corners (commissura palpebrarum lateralis) at the junction of the lids. *Cunningham (11) gives the nasal index of 23 Australians (native males) which “are ranged in the immediate vicinity of 94” with extremes of 79 and 104. This at once suggests a relationship between the Australian aboriginal and the Igorot. THE BENGUET IGOROTS. 439 Eye measurements. i = [ ] | | Group. Num Inner. | Outer. ye Eye width. 14 3. 60 9. 20 2.80 | Parisians 2.75 46 3.30 9. 20 3.10 | Belgians 3.00 44 3.35 9.05 2.85 | Chinese 3.20 "toa 3.40/ 9.10] 2.85 | Australians 3.34 | 10 3.30 8.80 2.75 | Negroes (Africa) 3.38 i a Topinard (6s). The eyes of the Bontoc are the narrowest and they are also more widely separated than those of the others(7). Those of the highland group are the widest and they are also the closest together, while the lowland is between the other two in eye width and the inter-eye distance. The eyes of the women are narrower than those of the men, but the same distance apart as the group to which they belong. The artistic conception of the Huropean eye is that it should be equal in width to the distance between the two eyes, and the artists add that the mouth should be one and one-half times the eye in width of opening. The Igorots have a smaller mouth and greater distance between the eyes than the artistic ideal for the Huropean (63). FACIAL ANGLE. The facial angle is determined directly with two brass bars bolted together at one end. One bar is placed in line with the external auditory meatus, the other with the glabella, and the apex of the triangle is opposite the point covering the junction of the nasal septum and the upper lip (subnasal point). This is not so accurate as the facial angle of the skull, and minor differences are not to be detected by this method, but it affords an approximate angle with ease and facility. The facial angle of the Bontoe Igorots is not measured, but that of the Highland and Lowland Igorots is given with that of the women and of the boys. Facial angle. Facial angle (per cent). Group. Nun ace Cephalic index. eer | | 65° to) 71° to} 76° to) 81° to} 86° to | | 70° | 75° | 80° | 85°} 90° ‘BOntOC === ees 0 0 Dolichocephalie____| 24 0 25 46 29) 0) Bighland pea 42} 78° Mesocephalic_______ 48 2 40 33 21 | Wo wlan dienes 17 | 779 Brachycephalic ____ 10 10 30 50 10 0 GROVE | 59} 779.5 | | | |eeVVON en eeaseen eames 10} 80° | ROVS!d Mae 2| 6° | Boys 6-10_------__- 3 | 78° | | Boys) 11-12_--------__ 5 | 76° | | Boys 13-14__---------- 5 | 79° | | Boys 15-16__-----___- 6 | 78° | } 440 _ BEAN. The Highland and Lowland Igorots have practically the same angle, while that of the women is greater. The angle decreases with age, for example: Four boys below the age of 10 have an angle above 80°, after this age it is below 80°. When compared with the cephalic index prognathism becomes evident among the brachycephalic, while the dolichocephalic are less prognathous and the mesocephalic are clearly mixed. The brachycephalic resemble the Negrito in their prognathism. MALAY AGAINST IGOROT. Four indices which are considered to be important in type differentia- tion may be compared with the same figures of the inhabitants of the Malay Peninsula(44). The cephalic index breads will be taken first, : length because it is the most important. Comparison of cephalic indices of Igorot and Malay. [eens y] | . Group. Index. | Group. | Index. Coes erie he Ea| | 3 M * ey =a ey aa | | | BONO Serena ene eee ee eal | iil | IBIS as} eens ene 77.1 Te bfeanibyavole tes ea | en di@e S| | MSM ON) Peewee renee reese eee eee 77.9 [Nexto wl airs cl pemesasleeeaeioeenlemn | 78.5 || Reine Senoi -__-_.------_______- | 78.5 The exactness with which the Bontoc and Blandas, the Highland and Semang, and the Lowland and “Reine Senoi” correspond is remarkable. The similarity between the Igorots and the inhabitants of the Malay Peninsula is so exact that if cephalic index is the criterion of type, the conclusion must be that they are identical types of people. The height index of the head is given next, because it is closely associated with the cephalic index. Comparison of height indices of Igorot and Malay. [PT . c ] ie | Group. Index. | Group. Index. | Bontoce2 cae hes ewe | 70.0 Blandas Highland __ To wlan Gi) ssese see eee The correlation of head height is almost reverse in its relation to the horizontal diameters (cephalic index). Could the low headed doli- chocephals be eliminated from the lowland group, the index would be raised higher than the highland, and the groups of Igorots would then THE BENGUET IGOROTS. 44] ‘correspond with those of the Malay Peninsula, excepting that the Igorot head is higher. The morphologic face index is correlated below: Comparison of the morphologic face indices of Igorot and Malay. Group. Index. | Group. | Index. RRS ES) ee! eto Se Ets el | | {5 Bontoc Se ae tae eee Peewee SASH |) AC as wees ee eens | 84.6 ls: Jahepow yO 2 LO |) ete avy os | (8225 Lowland sess eee eens 80.4 || Reine Senoi -__------_-_=_..-___- | 82.5 | BESI Sit tes es eees Cpe Se NS | 81.8 Martin’s groups have longer faces below the eyes than the Igorots. The lowland is more like the Reine Senoi than are the Bontoe and highland groups, which diverge from the other Malays. This may be explained on the assumption that the Bontoc and highland groups have greater Kuropean intermixture. Finally the nasal index is correlated as follows: Comparison of the nasal indices of Igorot and Malay. Group. Index. | Group. Index. | | i] BODOG! sat32 ese ees 97.6 | IBIATGaS eee ee eee Ne eae | 76.6 | labeaoveyool 2s 88x4al|bSeman pierre. = ea ne ee 83.5 owlandys:¢25so2 25 sees 95.0 || Reine Senoi _----_-_------------ | 86 | 1B CSS) enantio | 78.9 | | red ae DITO Pte ees oy eile | The disparity between the groups is marked. Again, if the type of dolichocephalic Igorots with very wide noses be eliminated, the disparity is diminished. The Malays of the inland part of the peninsula according to Martin are mesocephalic with one group brachycephalic; hypsi to orthocephalic ; brachyfacial to mesofacial; and mesorhinian to platyrhinian. The Igo- rots are mesocephalic and dolichocephalic; hypsicephalic; brachyfacial ; and platyrhine. They are also slightly prognathous. Vv. DESCRIPTIVE CHARACTERS. The skin of the Igorot is characteristically light brown, but the tint varies with individuals and it is different in different families (72,73). The influence of light and shade may be noticed; those who work in the sun are darker than those who serve in the house and the women and the children are lighter than the men. The whole family of one chief, including several young men and women who stay indoors a great deal is so light brown in color as to be classed as yellow. In a few individuals 442 BEAN. a tinge of red may be seen, or the face appears bronzed, some Igorots strikingly resembling the North American Indian. The coloring shows a trend towards lightness rather than the reverse, and this is manifested most strongly among the Bontocs. Skin color (per cent). : | Num- | Golden-| Light | = Dark | Group. ber. | brown. | brown. | Brown: Doe ms | BOD COC a see Soe ie tn cesT 125 eee | 30 | 60 5 | Highland 45 Sip veil (eer 13 | Lowland 35 2 | 8 | 83 8 Total 92 2) ies) emer 1 = i | The relative number of brown individuals increases in the Highland and reaches its limit in the Lowland. The lightest colored individuals are found in the highlands. The one golden-brown individual of the lowlands is a young man who for several years has been a servant in an American family, where he worked principally indoors and wore the regular Kuropean clothes of the Tropics. The hair is invariably black, straight, and coarse. A few individuals with wavy hair were observed, but not one of those measured had a noticeable wave in the hair. This is remarkable when one considers how closely the Igorot resembles the Negrito in other characters. I can account for the predominance of the straight hair in one way only— it is dominant to the kinky hair of the Negrito, and in the course of centuries the kink has disappeared leaving only an occasional trace, such as the few wavy-haired individuals I observed casually, and those noticed by Jenks(20) among the Bontoe Igorots(71). The wavy-haired individuals probably belong to the Senoi type of Martin. The brows of the Igorots are never so beetling, and the brow ridges never so prominent as among the Filipinos of the coasts and other parts of the Islands. However, there is a slight difference of the size of the superciliary ridges among the Igorots which may be presented in three groups, small, medium, and large. Brow ridges (per cent). | | Group. | Number. Small. | Medium. | Large. || SBom toe: 2 eo ee Bae ey a | 10 20 70 10 | 45 11 42 47 | 42 16 65 19 10 70 BAD) | pe THE BENGUET IGOROTS.. 443 The brows of the highland group stand out clearly, because they are larger than those of the other two groups, and the brows of the women are small, as is to be expected. The Igorot nose may be divided into three classes by the profile view— aquiline, straight, and australoid. With the side may be coupled the front view, in which two factors claim attention, the direction of the nostril openings, and the amount of flare to the ale of the nostrils. Hach of these characters has three qualities which may be combined with the three of the profile to make up three composite types. The aquiline nose has narrow nostrils that open downward and the nasal index is low. The straight nose has wider nostrils that open down- ward and forward, and the nose is compact without extremely flaring nostrils. The australoid nose has wide flaring nostrils that open almost forward and the nasal index is high, the nose extremely platyrhine. Types of nose (per cent). Group. Number. |Aquiline.| Straight. | Australoid. BONtOC Sheen sewn es ees 5 0 80 20 PEL hh 8 eee Sa eee ee 46 16 28 56 TsO Wien dieses Sate eee oe ease es 32 6 22 72 | Totalie sso vle see ee ee 83 12 30 58 SW OMe Tacit en ben esau 10 | 10 70 20 Of these three types, the aquiline is found most frequently among the Highland Igorots, the straight among the Bontoc, and the australoid among the Lowland. The nose of the women is usually straight or australoid. [Plates II, III, and IV.] HEAD OUTLINES. The head outlines are treated as composites in groups, according to cephalic index and by locality. Only the sagittal outlines are utilized because they are more accurate than the other, and illustrate more dis- tinctive differences. The composites are made by drawing each outline on transparent paper with the mid-point of the line which connects the glabella and the occipital tubercle, as well as the line itself, superimposed upon the same point and line for each drawing. After all the outlines of one group are drawn in this way the heaviest line is eppnarucee as the com- posite on another sheet of paper. The composites grouped according to cephalic index indicate what is to be expected from yon Torék’s classification, namely, the dolichoce- 444 BEAN. phalic heads are the largest and the brachycephalic the smallest, while the mesocephalic are intermediate in size. (Fig. 3.) The composite \ | I Fic. 3.—Composite sagittal outlines of 104 Igorots: 41 dolichocephalic, largest outline ; 43 mesocephalic, intermediate outline; 18 brachycephalic, smallest outline. curves of the three groups are similar. The forehead of the brachyce- phalic protrudes slightly and the occipito-parietal region is somewhat flattened. When the composite dolichocephalic head outline of the Igorots is compared with a similar outline from an equal number of negroes I measured in Baltimore at the Johns Hopkins Hospital Dispensary in 1906, the data being as yet un- published, and an equal number of white students of the University of Michigan 1 measured at Ann Arbor in 1905 to 1907(4), some striking differences may be seen. (Fig. 4.) The head of the Igorot is the tallest and shortest of the three, y Ss 4 Se Ee j \ \\ \ | \.\ \a al \, \ || \ |: \\ | ' seh Nai aes ee aes BE fete! oe ME PaIRES Fic. 4.—Composite sagittal outlines of the dolichocephalic heads of Igorots, negroes, and white students: Igorot, the short, high outline; white, the long, low outline; negro, the broken outline. whereas that of the white student is the lowest and longest. The forehead of the negro is low and receding, while that of the Igorot and white student are high and prominent. The head region immediately above the somasthetic area THE BENGUET IGOROTS. 445 of the brain is prominent in the Igorot and in the negro, but not in the white student. The white student has a relatively large frontal region, the negro has a relatively large body sense and motor region, while the Igorot has both. The Igorot represents a protomorph, or a mixture, while the negro and the white student represent specialized products of evolution, or definite types. The Igorot contains elements similar to each of the others, at least this is true of the dolichocephalic. : The brachycephalic head outlines reveal somewhat different characteristics. (Fig. 5.) The white student is again the longest, but it is also the tallest, the Fic. 5.—Composite sagittal outlines of the brachycephalic Igorots, brachycephalic negroes, and brachycephalic white students: Igorot, the inner solid outline; white, the outer solid outline; negro, the broken line. Igorot has a rounded outline with full, high forehead and the negro has 2 bombé forehead high in the frontal region. There are only 6 brachycephalic negro head outlines and the composite for that reason is not a representative one. The mesocephalic head outlines represent more defirfitely than the dolicho- cephalic the important differences denoted by the latter. (Fig. 6.) The Igorot Fic. 6.—Composite sagittal outlines of the mesocephalic Igorots, mesocephalic negroes, and mesocephalic white students: Igorot, the broken outline; white, the long, solid outline; negro, the short, solid outline. 78322 3 446 BEAN. head outline is an exact blend of the white student and the negro, except that it is shorter and slightly higher. The white student’s head is relatively large frontally, the negro’s is relatively large parietally, and the Igorot’s is relatively well developed throughout. The sagittal outlines of the 10 Igorct women of Atoc when compared with similar outlines of 10 women students at the University of Michigan selected with the same cephalic index show great dissimilarity. (Fig. 7.) The white _ Fic. 7.—Composite sagittal outlines of the heads of 10 Igorot women and 10 American college women: Igorot, short, high outline; American, long, low outline. student head is longer, the Igorot head is shorter and higher, and both are depressed beneath the occipital and frontal regions. The somesthetic region of the Igorot is protuberant, while that of the white student is unobtrusive. The small number of individuals and the difficulty with the long hair of the women vitiate the records somewhat. but the differences in height and length are of the same nature as in the men, although intensified in the women. The distinguishing differences between the Igorot and the American student are the length and height of the head. The Igorot’s is short and high, the American student’s is long and low. The Igorots, male and female, carry heavy burdens by straps across the top of the head, which may influence the height of the head from compensatory hypertrophy following the continual stimulus of great pressure. The generous muscular development of the Igorot may also have some influence in heightening the head, by increasing the size of the somesthetic area of the brain. The head height is a racial trait, as well, which may be inferred by comparing the sagittal outlines of the three groups of Igorots. (Fig 8.) The Highland Tgorots have longer, higher heads than the Lowland Igo- rots. The Bontoe head is longer than the others, but not so high as the Highland. This is an additional differentiating fact between the three groups, and again the Bontoc is more nearly like the white, whereas the Lowland is less so than the others. THE BENGUET IGOROTS. 447 Fic. 8.—Composite sagittal outlines of the three groups of Igorots: Bontoc, the broken outline ; Highland, the large solid outline; Lowland, the small solid outline. EARS. The ear of the Igorot is a most typical feature and a true racial char- acter. Not all the ears are alike, indeed there are at least three well- defined forms, and many yariations of the three. he typical Igorot ear is found oftener than any other kind, and its frequent presence merits a special description and portrayal by photograph. (Plate V.) The typical Igorot ear is large and long and somewhat rectanguar in shape. The superior border of the helix is smooth, thin, gracefully rounded, and the posterior border is straight. The anthelx circum- _ scribes the concha in the shape of a large oval with its apex at the inciswra intertragica. The lobule is square and flat, the inferior border usually joming the cheek at right angles. The ear does not stand out from the head, neither is it pressed close to it, but occupies an intermediate place, and is beautiful and graceful in both form and position. (Plate VI.) There is not a line or character about the Igorot ear to relate it with the anthropoid apes nor with any of the primitive people of the world, so far as I am able to judge. It is not like the Negrito ear, which is short and round, the helix of which passes horizontally back- ward from the superior end of the base, the anthelix forming a roll that often gives the ear the appearance of having a double helix, and the lobule is round or pointed. (Plates VII and VIII.) It has none of the characteristics of the Australian aboriginal ear(1!) which is similar in many ways to the anthropoid. Darwin’s tubercle is present more frequently in the men than in the women, which is true of Europeans (52). I have seen ears resembling the Igorots on Spaniards, Englishmen, and 448 BHAN. Americans, but 1 shall discuss that subject fully in a forthcoming article on Filipino ears. The ear is a European one, and characteristic of one of the finer types of Europeans. The other types of ears among the Igorots resemble the Negrito, and the Malay or Chinese ear without lobule. Three types of ears may be distinguished, which are distributed as follows: Types of ears. | ae | Oval, no | Group. Number. | Typical. Topule® Round, pest W=3Bon toc an De eee 12 9 2 1 Mages Frsaub yr ts acetate mee een reee eee 45 29 | 12 | 4 | Toy learn yee Oi eee ee | 34 20 | 6 8 | for Ey yeep tr eee egan ees SS ss 91 58 | 20 13 | | Women 2i23 S55 Se ee ees 10 9 1 0 | | BoySat2e 22 ee 27 20 3 The Bontoe Igorots have a relatively larger number of typical ears than the other groups, although the Highland Igorots have almost as great a relative number, but the Lowland have the least, and also a greater number of round ears (Negrito?). The ear index of Topinard (65) is useful in differentiating the types, but Schwalbe’s morphological index was not utilized. The ear index greatest breadth x 100 is ehe greatest length Ear measurements. Group. oe Length| ee | Dini Teale (MeSH). | oT ea re | | | Bontoc sae 12] 33,1] 57.2) 57.8 | Type. tae Index. | Highland __________ 46| 31.9] 59.3] 53.8 || | = HP oy A wave See 44 | 31.4 56.7 55.3 || Ruropeans__-______- 8| 54.0 | Total | tof) 31.8! 57.9) 55.0 | Melanesians_______. 8) 59.5 | Women___----__- wear 10 | 28.8 | 49.7 57.9 || Polynesians -----_- 3} 60.0 | | | Negroes (African) __ 13] 61.2 Sct CA eae te Bees IT. Brats A [esate <8 | The Highland Igorots have the longest ears and the lowest index, with the Bontoe second and the Lowland third in ear length, but the Lowland index is less than the Bontoc. ‘The ear index is not an absolutely reliable indication of ear type, but with the aid of descriptions it is serviceable. The index of the typical Igorot ear is low because the ear is long and not round. The ear should be one of the best marks to determine the nature of heredity, because it is not subject to sexual selection in the way that other features such as the eyes or nose may be, THE BENGUET IGOROTS. 449 and there is no reason to believe that natural selection would affect it. For the same reasons this feature should be one of the best marks to determine racial purity. The ear marks of a people may be significant. CORRELATIONS. The correlation of cephalic dex and stature is determined by averages and percentages. The average stature of the dolichocephalic Igorots is 157.1 centimeters, that of the mesocephalic is 155.2, whereas the stature of the brachycephalic is only 152.2 centimeters. Correlation of cephalic index and stature (per cent). | Stature | Stature | Stature | Tindex | below |150 to 160,above 160) Eee /150 centi-. centi- | centi- | | meters. | meters. | meters. Dolichocephalic=== = eee 16.6 64.2 Bi) Te. | | | lee Miesocep hall ic tases meese nee eerie eee ia eee 20.0 | 62.5 17.5 laeBrachy Cephalic resem tenia ee eee 31.8 54.5 13.7 | There is a greater proportion of comparatively tall individuals among the dolichocephalic Igorots, and a greater of small individuals among the brachycephalic than among the mesocephalic, but the difference between the mesocephalic and brachycephalic is greater than that hetween the mesocephalic and dolichocephalic. A larger per cent of each index is found between 150 and 160 centimeters, which is to be expected in a much mixed, endogamous people. However, it is in the extremes that aboriginal types are to be searched for, and it is the extremes where the differences are greatest. The correlation of cephalic index and relative arm length is not so great as the correlation of cephalic index and height, but it is in the same direction. The long head and the tall height are parallel and so are the long head and the relatively long arm. However, the correlation is shght. Correlation of cephalic index and relative arm length (per cent). Below | Between) above, Groups. 43.0 centi- “sts conti 45.5 centi-| meters. | heters, | Meters. | | ae ae ee | Dolichoceph alli chasse sae ae ee 14.2 64.4 21.4 | Brachycephalic 26.3 52.3 21. | Mesocephalic 31.0 54.8 14.2 450. BEAN. The average relative arm length is 44.3 for the dolichocephalic, 44.2 for the brachycephalic, and 43.9 for the mesocephalic. In groups above 45.5 there are 21.4 per cent dolichocephalic, 21 per cent brachycephalic, and 14.2 per cent mesocephalic. In groups below 43.0 there are 14.2 dolichocephalic, 26.3 brachycephalic, and 31 mesocephalic. The dolichocephalic have relatively longer arms than the mesocephalic, while the brachycephalic have an intermediate relative arm length. Correlation of stature and relative arm length (per cent). | Below 43.0 to | Above Stature. 43.0 centi-/45.5 centi-|45.5 centi- meters. | meters. | meters. Above! 160{centimeters =2222 sae 9 82 9 150 to 160 centimeters —__.------2----------- eeosts 53.4| 28.3 Below 50\centimeters| 22s ae eae | 38 47.6 | 14.4 | There is a progressive increase of relatively short arms (below 43) from absolute tallness to absolute smallness, and a progressive increase of arms of relatively intermediate length (43 to 45.5) in the opposite direction, while the number of long arms (above 45.5) increases from tallness, through smallness, to medium size in stature. Continuing the analysis of cephalic index combined with stature and relative arm length a table is presented as follows: Analysis of cephalic index combined with stature and relative arm length (per cent). Below 43.0 to Above Stature and group. 43 0 centi-/45.5 centi-|45.5 centi-| meters. | meters. | meters. | Above 160 centimeters: Dolichocephalic eee ten en a eamre eee! 0 100.0 0.0 | Mesocephalics. o> te eee 12.5 62.5 25.0 Brachycephalic eee ee 20.0 80.0 0.0 150 to 160 centimeters: Dolichocephalic= sss eae ae ane en ee 14.8 55. 6 29.6 Mesocephalics== ss Ste eee 30. 4 56. 6 13.0 Brachy cephalic eee 30.0 40.0 30.0 Below 150 centimeters: Dolichocephalic=== eee 33.3 50. 0 16.6 | Mesocephalicss 2) sa. Se een | 50. 0 40.0 10.0 Brachycephalic== ee eee | 20.0 60.0 20.0 The Igorots may be divided into four groups by the above correlations: 1. Tall dolichocephalic Igorots with long arms. 2. Small dolichocephalic Igorots with short arms. 3. Mixed mesocephalic Igorots. 4. Brachycephalic Igorots with intermediate arm length. The correlations and differences suggest that three steps of racial mixture preceded present conditions. First, a small dolichocephalic THE BENGUET IGOROTS. 451 people with relatively short arms and a brachycephalic people mingled and partly fused. They were then joined by a tall, dolichocephalic, long- armed people already partly fused with the brachycephalic, and sub- sequent fusion was again altered by contact with the brachycephalic people. The last contact was quite recent and the brachycephalic people are more distinct as a type than either the tall dolichocephalic or the small dolichocephalic, and they are also present in greater number. VI. SOMATOLOGIC RACE TYPES. Stratz(56) divides mankind into three groups, protomorphs or nature folk, archimorphs or highly differentiated peoples, and metamorphs or mixed races. These may be used in connection with the canon of Fritsch and the artistic modulus(24) as comparative standards. The canon of Fritsch takes as its standard the length of the vertebral column and the other body measurements are compared with this(57). The length of the vertebral column is equal to the distance from the symphysis pubis to the nasal spine. With this basis, photometry may be made an adjunct of anthropometry when interpreting the length relations of the body parts. The artistic modulus is the total head height from chin to vertex, and it is used in relation to stature. The modulus of Geyer, which is the stature equal to 8 total head heights, is the artistic ideal for the European. With this explanation the following classification is given. The protomorphs comprise the Australian, Papuan, Hottentot, American Indian, Eskimo, Philippine Negrito, and the Pigmy. of Africa. The archimorphs are the leukoderm or white, the melanoderm or black and the xanthoderm or yellow men. The metamorphs are mixtures of the other groups, and are found along the zones between the black, white, and yellow races; in northern Africa, eastern and southern Asia and in the islands of the Pacific. The protomorphs are short in stature with relatively long total head height, which is in the lower face and not in the cranium, and their arms are relatively long. They conform to the canon of Fritsch except in the relative length of arm, and to the artistic canon except in the relatively large head. The melanoderms are relatively short in stature, long in arm, and short in upper head height, nasal spine to vertex. The xanthoderms are relative short in stature, in length of leg and in upper head: height. A slight departure from this may be noted in the females of each group. ALBNIQOT S'la | 9°08 ta 4 P°8z TESG Sealine praia, | Vribmesryes lien ap ast CKca GS £08 o's 60805 | ances | hae 8°1% ORG tell ie tae ik AIBNUBL sumer | ume | ‘ume | cumu | emer | enor | ume | sermon | sermon | “ror | nu | urn | oro | cn | oro | cu | “tor | UU “LUNI | -EXBIT | -TOLIN aur | xe | Tou | -pxen | TOU | XB | “TUN | “EXO | “TON | TXB | “TUT | -EXBIN | “TOL 1X8 ee ———— = — 1 = — "QUOW ‘OSBIBD) ‘uRnjng ‘uByided *07BQ'BJOD “OBAB "RSUROGUIBZ, “BlOqBST ‘olor’ ‘oBsIINg ‘LOGE Siwaunp suoynjs opunpw yt lof Sappsbyjuas soawhap wy ‘saimposadiuay Ununiue wpa pup wununEDUL Upaul Huynoys 3)QD,1, GEOLOGIG RECONNAISSANCE OF MINDANAO AND SULU. 481 The region around Camp Keithley, owing to its altitude is much cooler than the low country and the climate is correspondingly invigorating. However, at certain times of the year, particularly in December and January these posts are said to be very disagreeable, as they are cold and raw like the New England coast of the United States in the spring time. Typhoons are said not to occur in the latitude of Mindanao and the Sulu Islands. The evident reason for this is that the cyclonic storms, which have their origin in the Pacific are formed in a latitude much north of that of Mindanao and as they pass westward they are constantly curving to the north, partly owing to the original, clockwise movement of cyclones north of the equator and partly because of their approach to the continent of Asia. According to Father Algué, Director of the Philip- pine Weather Bureau, a few cyclones form in the Sulu Sea, but these attain no great importance either in frequency or in intensity. Plate XXIX of Father Algué’s “Cyclones of the Far Hast” *° shows the mean trajectories of cyclones which pass over or near the Archipelago. Tt is very interesting in that it reveals how very generally Mindanao and the Sulu group escape these destructive storms. This fact is of the first importance in view of damage which such storms might inflict on crops, particularly on hemp which grows to heights varying from 10 to 18 feet, and because of the relative immunity from danger to vessels, such as interisland trading ships, Moro vintas and pearling boats. VI. NARRATIVE OF THE BPXPEDITION. Zamboanga, the first point visited by me, is situated about 3 miles from the nearest foothills at the edge of a flat plain of considerable area at the foot of the long, narrow peninsula of the same name. (See map, Plate I.) To the east is a long stretch of salt-water marsh and in its rear is a scarcely less elevated tract which is taken up with paddy fields. If the Tumaga River had kept its initial direction, it would cut through the heart of the city, as it is, it curves to the east and enters the sea op- posite Sakol Island. The substructure of this plain is coral, the super- structure, silt and coarse detrital material from the hills to the north. Zamboanga is essentially a “gate city” and a study of the map will show its central, commanding position with reference to steamship routes. In fact this is the main feature controlling its location. It is not situated on a large river by which communication can be maintained with the interior and for this reason its position is not favorable as is that of Manila, which is on a plain on the coast and at the same time on the banks of a large stream which taps a great stretch of the interior. Cota- bato, on the Rio Grande de Mindanao, is also favorably located and it will probably expand when the immense possibilities of the country to which it holds the key are understood. ® Algué, José: The Cyclones of the Far Hast, Bureau of Public Printing, Manila, 1904. 482 SMITH. The initial reconnaissance which I undertook was to Boalon, some 10 or 12 kilometers northeast of Zamboanga. Here the transportation by wagon was left and a trail taken which led up an abrupt hill a little beyond which point we entered the forest. Between Boalon and this hill I found some float limestone with fragments of Orbitoides, which probably are identical with the material Richthofen ** encountered so many years before. This is practically all he contributed to the geology of this region, but this is not suprising when the attitude of the natives at that time is considered. We continued in the forest for about three days, obtaining absolutely no view of the country farther than 50 yards from the trail, until we reached a log cabin about 30 miles north of Zamboanga on the Tumaga River.** The country rock in this region is a much decomposed schist, with a considerable thickness of stiff, yellow, clay overburden. (Plate III.) Quartz pebbles and bowlders are plentiful in the clay; the pebbles come from quartz stringers in the schist. The large bowlders clearly indicate large veins, but we were not so fortunate as to encounter any of the latter. The clay contains a small amount of gold which the prospectors had recovered by sluicing. Few people were encountered in this forest. On our return after three days’ stay we followed the river for per- haps 25 kilometers. At times we came upon box cations (some of which we might have swum through), but usually we took the high trail which went along the steep side of the cliff, at times 30 meters above the water and rocks. The way was extremely difficult; the sharp river rocks, the sharp-edged schists and the leeches began to tell on our carriers, so that we finally took a trail which led out of the river and after a climb up the side of the gorge we regained the old path which we followed to a hemp plantation at the edge of the timber and by mid-day arrived in Zamboanga. VICINITY OF SAN RAMON. My next reconnaissance was in the vicinity of the San Ramon Farm, controlled by the Moro Province. This excursion was for the purpose of an examination of the mountains which rise abruptly back of the narrow coastal plain. The formation in this place is entirely volcanic, the rich, disintegrated débris spreading out upon the coastal plain and producing a very fertile soil. Some of the rock is highly pyritized and may carry more or less gold. nN In the streams I saw bowlders of andesite which sometimes contained large fragments of schist, torn off and caught up in the molten rock as it * Richthofen, F. yon: Vorkommen der Nummulitenformation in den Philippinen, Ztschr. d. deutsch. geol. Ges. (1862), 14, 357. 4 A sketch map of the trail was made, but it shows little beyond the path. GEOLOGIG RECONNAISSANCE OF MINDANAO AND SuLU. 483 poured out over the surface. This schist is identical with that found in ‘the gorge of the Tumaga River. I had observed terraces along the shore farther to the north in the neighborhood of Dapitan and therefore looked for some signs of elevation here. I did find one fairly well preserved terrace a few miles to the north of San Ramon, but the streams have cut through it in so many places that only an especially trained eye can see it. This terrace is perhaps 6 meters above the mean tide level. (Plate VII.) The mountains in the Zamboanga Peninsula were once covered with a mantle of limestone, but little of the latter remains, a few large bowlders in the streams being all that we could find. This coastal strip on which San Ramon is located, disappears to the north at Patalun Point, but it widens regularly to the south and is every- where taken up with coconut culture. The long stretch of sandy littoral from San Ramon to Zamboanga is especially adapted to the culture of the coconut palm. This peninsula in regard to its population might be divided into the following zones: 1. The hill or forestal zone occupied by Subanuns, a wild and primitive people. 2. The intermediate or rice zone by Filipinos, mostly Visayans. 3. The coastal plain by Chinese and Americans. 4, The beach zone by the Moros (littoral zone). COAL MINES AT SIBUGUEY. Following the reconnaissances outlined above I went to Sibuguey Bay, an all-night run by Constabulary vessel to the northeast of Zamboanga. I was accompanied on this trip by Colonel W. C. Taylor, then in com- mand of the Fifth Constabulary District. We anchored about a half mile offshore as we did not know the exact configuration of the reefs, this coast being but incompletely charted. Our road to the coal measures at Sibuguey was first by boat up the Siay River, the banks of which for some distance from the mouth are lined with mangrove swamps, we taking a turn through an opening in the right bank and following an estuary until noon, when we landed and followed a trail over a low hill to the house of the datw of this region, Lukas, a Subanun. (See Plate VIII.) We left this place early in the afternoon and after a yery trying mareh through mud and over hills, we reached the site of the coal workings on the Sibuguey River. The old, abandoned tunnels of the coal mines are - hidden by the underbrush. I could obtain but little idea of the condition of the seams, excepting that they are tilted and dip to the southeast. They can be worked with the mine mouth probably not over 100 meters from the Sibuguey River, which is large enough at this point, 21 kilo- meters from the mouth, to allow small launches and lighters to pass up and down to take on coal. 484 SMITH. THE SULU ARCHIPELAGO, The Sulu Archipelago is practically unknown from a geologic and physiographic point of view, so that the information gained in this expedi- tion, although meager, is at least new. Dr. Becker refers in four lines of his report to rocks on Marongas Island just across from the town of Jolo. Other than this there are no geological notes. Some eruptions of the year 1614 are described by the resident Jesuits but only in a crude way. It has been known for a long time from the reports of ship captains and travelers that the archipelago is largely volcanic. This great group of islands extends for 335 kilometers southwest of Zamboanga. It is about 120 kilometers wide and contains hundreds of islands and rocky shoals. The most considerable of the islands are Basilan, Sulu, Siasi, and Tawi-Tawi and although Basilan is the largest, Sulu is of far greater commercial and historic interest. The water is nowhere of great depth within the confines of this ar- chipelago, but it is separated both from Borneo and Mindanao by deep straits. To reconstruct a large island out of this swarm of small ones which may or may not have formed a continuous bridge from Borneo to Zamboanga, and which would since have been disrupted and partly sub- merged, is not a great tax upon the imagination. The evidence we have points in this direction. The Sulu group, like so many other oceanic islands, is either of vol- canie or coral formation. I saw very little sedimentary material on any of the islands and where any such was exposed, it was usually at a point where erosion had removed the lava capping. I do not know of any marked yoleanic activity in the Sulu Archipelago at the present time. although hot springs in old craters are reported on Cagayan Sulu. I have also been informed that there are hot springs and solfataras at SvVit Lake on Sulu. This lake occupies an old crater. I have visited neither of these places. An eruption at a point near Jolo, not named, is reported to have occurred on January 4, 1641.27° As there is some confusion of names in this report I do not attach much importance to the account. As far as I can learn, no accurate scientific notes were taken at the time. The only remaining reference I find regarding the geography or geology of the Sulu Archipelago is a note by Becker :1° “In the Jolo Archipelago, the charts indicate several well-developed atolls, such as Simonul Island (latitude 4° 52’, longitude 119° 50’), as well as several in the Tapul group (latitude 5° 30’). The charts of this region also show innumer- able coral reefs, which are bare at low tide and must therefore have been uplifted.” * Baranera, Francisco X.: Compendio de geografia de las Islas Filipinas. Marianas, Jolo y Carolinas. 3r ed. Manila, 1892. ** Geology of the Philippine Islands, 562. GEOLOGIG RECONNAISSANCE OF MINDANAO AND SULU. 485 I myself have seen some coral islands near Sulu, in the Pangutaran group; some of these are atolls and others were formerly lagoons that have dried up because of the elevation of the whole mass. While nearly all the elevations are extinct or dormant craters, there are no sharp, jagged profiles, but instead, most graceful curves. There are nearly fifty of these cones on the Island of Sulu, some still high and symmetrical, others irregular and worn down to mere stumps. (See IPL IW.) BUD DAJO. We first visited the now historic crater of Bud Dajo," the wooded cone of which rises from the plain back of Jolo. One afternoon of brisk ridmg on horseback is necessary to reach the point where the very steep climb begins at 300 meters’ altitude; from here to the ex- tinct crater is a further elevation of 580 meters. The climb is a short one, but it is the most strenuous I remember ever to have made. Formerly there existed a community on this mountain having all the necessaries of life about them; a complete village with dwellings and a mosque; springs, gardens, and both shade and fruit trees, all within an extinct crater. The Moros added trenches and cottas to the natural walls of the village and long bamboos, in the hollows of which were concealed lantakas (brass cannon) were placed along them. (See Plate X.) Bud Dajo is formed of scoriaceous basalt and lapilli, but has not been in eruption at least within the last three hundred years, or if it has, there is no record of the event. Large basaltic bowlders from this mountain are strewn over the slopes and the plain down to the very edge of the town of Jolo. An excellent view of a large part of the island can be had from the highest point on the walls of the crater. From this point smaller craters are visible and it is not impossible that renewed energy may at some future time be manifested at one or more of the many foci and a considerable destruction of lives and property be the result. The fact that these craters appear extinct is no argument against future activity. FURTHER RECONNAISSANCE OF SULU. On Monday, October 14, Lieutenant Caffery and J, with an escort of five men, began an expedition to Maymbung, on the opposite side of the island. he trail led past Asturias, the former residence of the Sultan of Sulu, but now the site of infantry barracks, and on over a low divide of about 300 meters’ altitude, between Bud Agad and Bud Pula. The soil is of a rich red color, giving promise of unusual richness. Large fields of tapioca, which is the main agricultural product of the * Bud is the Sulu term for mountain. As it is generally used by the military authorities, it is retained in this description. 486 SMITH. Sulu Islands, appear on either side of the trial and Moro dwellings, with several small haystacks near by could be seen from time to time. How- ever, large tracts of fallow land exist along the route. The soil everywhere is the same, for a blanket of basalt apparently lies over the whole island. At Maden Patung, about a mile and a half from the Sultan’s house at Maymbung, are some outcrops of tuff, the only sedimentary formation I saw while on the Island of Sulu. We reached Maymbung late in the afternoon and the next day returned to Jolo by the same route we had come by, our stay bemg cut short by the consideration that a geological reconnaissance conducted under guard in a very unsettled country does not warrant the expense and the addi- tional detail of men. “Such work is really more exciting than profitable.” Several short excursions in the vicinity of Jolo were made for the purpose of finding water-bearing strata, but in this respect the result was disappointing. However, some splendid examples of old, worn-down craters were seen. Several low, circular and apparently flat-topped hills lie at a distance of 3 to 5 kilometers southwest of Asturias. They very much resemble overturned saucers. The tops of these hills usually show a more or less marked depression, a remnant of the old crater, and two of these were inhabited by several families, with substantial houses and well-kept gardens. These people live in such situations, not so much because the soil is particularly rich, because it would be hard to find soil more fertile than that on the lower volcanic slopes, but undoubtedly because of the protection afforded by the hills, the comparative difficulty of access and the excellent lookout over all approaching trails. There is usually some water either in the central depression or at the bases of these volcanic mesas. OTHER ISLANDS OF THE SULU GROUP. I returned to Zamboanga after this brief visit to Sulu and reshipped on a small Constabulary paymaster-boat for the more distant islands of the Sulu group. The first stopping place was at the Island of Bongao. Tawi-Tawi was not visited, such observations as were possible being made while sailing near to the coast. It is not a very rugged island, everywhere showing gentle curves. Bongao is a small village and Constabulary station on the island of that name, separated by a narrow channel from the southwestern end of Tawi-Tawi. Coral reefs are found everywhere in these waters, so that great care in navigation is necessary. Mount Vigia, visible from the dock (see Plate XIV), is a mass of very resistant conglomerate, 370 meters high, and on a clear day the low coast of Borneo can be seen from this peak. Some raised beaches exist in this vicinity and a number of fossils of recent age, clearly Pleistocene, were procured. From Bongao we navigated through a labyrinth of islands along chan- nels so narrow as to make it almost possible to lean over the side of the GEOLOGIG RECONNAISSANCE OF MINDANAO AND SULU. 487 boat and touch the branches of the trees, and we finally anchored in the narrow straits between Siasi and Lapac Islands. We had but one or two hours of daylight at this point, but a short excursion inland gave us a fairly good idea of the geology and soil of Siasi. The soil, as in Sulu, is a rich, red volcanic material, and the underlying rock, wherever I saw it, was andesite or basalt, which is frequently difficult to classify exactly, because of the weathering to which it has been subjected. Very little timber is seen on this island, at least not on the side at which we touched, and there are no large streams. However, a dense growth of cogon grass prevents serious damage from erosion. If this grass were not present, loss would surely result owing to the lack of forest. On the other hand cogon, as in other parts of the islands, is a serious menace to agriculture. It is usually the custom of the natives annually to burn off this grass, but this method only affords temporary relief. A better way and one which is being practiced with success in many localities is to plow the cogon under for two or three seasons, when the roots rot and not only is the grass killed, but the soil is further enriched. - The next point visited was the large and geologically little known Island of Basilan. A portion of this island was occupied by the Spanish government, which had a small naval station at Isabela on the north coast. This has been abandoned since American occupation and the place has consequently fallen into neglect and decay. But little geographical exploration has been done in Basilan. Mr. Dean C. Worcester and his party visited it about the year 1892, and the following is taken from his account.® “Isabela, the capital of Basilan, is a small place of less than 1,000 souls. The only Spaniards there are the officials and the Jesuit priest. The town is on high ground, which slopes sharply down to the edge of the channel separating Basilan from the little island called Malamaui. This channel, although extremely narrow, is very deep, and large vessels can come close inshore. Tremendous currents rush through it with the ebb and flow of the tides. “Tsabela is a supply station for gunboats, the coal yard and magazines being located in Malamaui, just across from the town. To defend the important stores which they contain there is only a ridiculous old limestone fort on a neighboring hill, armed with two or three antiquated smoothbore cannon, and garrisoned by a few marines. “The Moros of Basilan, locally called Yacans, have always borne a bad reputation, but at the time of our visit they were held in check by a remarkable man known as Datu (Chief) Pedro.” A picture of the fort mentioned above is shown on Plate XV. Only two short trips were made into the country back of Isabela; the same basaltic flows and rich red soils exist here as in Sulu and Siasi. Vulcanism does not appear to be as recent in Basilan as in Sulu. At no place did I find that the streams had cut through the lava capping and , exposed the sediments which I feel sure lie beneath. #8 Worcester, Dean C.: The Philippine Islands and their People. Macmillan, (1901), 144. 488 SMITH. LAKE LANAO AND VICINITY. All of the work in the region of Zamboanga and the Sulu Archipelago which it was at all feasible to undertake at this time having been com- pleted, we left Zamboanga, November 6, for Overton. As our vessel kept close to the coast, I was able to make some notes which throw considerable light on the geologic changes now going on. ‘The west coast of Mindanao has very certainly risen in comparatively recent times. Near Point Blanca on the northwestern part of the coast I saw a fine example of a raised delta, the elevation amounting to at least 10 meters. The char- acteristic structure of the delta was clearly revealed by the extensive marine erosion which had taken place. ‘There were also many fine terraces shown along this coast and their existence supports the other evidence. The weather compelled us to run into a little cove near the point just off Dapitan. Of all the many inlets along the coasts of these islands I believe this to be one of the prettiest and most secure. No sign of an entrance can be seen at less than a kilometer away and certainly this point would be too*obscure to pick up at night. We went through a channel not over 45 meters wide between walls which in the darkness I took to be limestone, and emerged into a splendid basin with water as clear and placid as a mountain Jake and with high walls on nearly all sides. By noon of the next day we anchored off the little stone fort at the entrance of Panguil Bay, which is in the extreme southwest corner of the much larger Bay of [ligan. The most conspicuous object at this place is Mount Malindang, an extinct voleano close to 2,700 meters in eleva- tion lying to the west. Material from the slopes of this mountain is basaltic as I discovered by going up Panguil Bay in a banca in company with Lieutenant Lattamore, Philippines Constabulary, and a detachment of soldiers, landing at several points to enable me to go far enough inland to examine the rocks, as there are no outcrops on the coast. Mount Malindang is an old crater the rim of which is broken down on the side toward Misamis. It is for the most part covered with a luxuriant mantle of timber forest; the soil on its slopes is of a rich red and is undoubtedly very fertile. : A number of Visayan colonies exist on the west side of Panguil Bay, but all the country to the east is Moro. We left this point on the morning of November 11 and reached Camp Overton at a little after noon. Mr. Ickis was to join me at this point, but as he was detained by quarantine, Lieutenant Caffery and I went forward over the military road to Camp Keithley (745 meters) where the climate is much cooler than in the coast towns. There is very little coastal plain in the region of Camp Overton, the hills rising so abruptly that the road has to wind back and forth in order to make the ascent. The first part passes through raised coral reefs, in GEOLOGIC RECONNAISSANCE OF MINDANAO AND SULU. 489 which the species are for the most part identical with those growing in the sea below, but within about 300 meters the road cuts through basalt, and continues in this formation until Malabang on the southwest coast of this portion of the island is reached. Maria Christina Falls are situated but a short distance off the main road, a few kilometers out from Overton. Here the swift waters of the Agus River, which drains Lake Lanao, fall over a cliff 58 meters high and continue to the sea through a narrow gorge. It has been estimated that sufficient power can be developed by these falls by means of turbines to run electric freight and passenger trains from Camp Overton to Camp Keithley and then around the lake and down to Malabang. Furthermore; the power from this and the Mataling Falls together should also be able to furnish electric light for a dozen towns and camps along this route. The rock in the upper portion of the section at the waterfall is a hard, rather structureless basalt; below this comes a more or less loose volcanic conglomerate, or better, agglomerate, the geologic structure giving the most favorable conditions for fall formation. Maria Christina has about the same height and yolume of water as the better known Majayjay Falls of Luzon. The road, very soon after the fork to the waterfall, leaves the rather heavy timber. From here on it ascends a long, gradually sloping, quite open and rolling plain, resembling the western prairie of the United States. The Agus flows in a broad valley with gently sloping sides at Numun- gan and while at this point it has a fairly rapid current, it gives no intimation whatever of the terrible plunge a few miles farther on. A party of engineers is stationed at Pantar some distance beyond this point to look after the roads and bridges, and we spent two days here to examine the cuts along the road and river bank. Basalt is still the country rock here, but it has on top an extraordinarily thick mantle of weathered material full of basalt bowlders, and both in constitution and topography this simulates glacial morainal material. From Pantar the road runs fairly straight for seven to nine kilometers across open rolling country to the “Keithley escarpment.” Beyond this escarpment lies Lake Lanao. The road continues almost due south to the foot of this great wall, then turns practically due east and, keeping nearly parallel with it, climbs gradually to the top. From here it runs down a long, easy grade to the margin of the lake. This escarpment is very striking, and is made up from top to bottom, as far as can be seen from its cuts in the road, of loose material, unsorted and with apparently no definite structure, forming a wall 155 meters high. A simple explanation of this phenomenon is not easy to find. To the left, when facing toward Camp Keithley from the top of the embankment, rises the dark, heavily wooded mass of “Sacred Mountain” 490 ‘SMITH. some 300 meters higher. In the middle distance stands a prominent, grass-covered bump known as “Signal Hill” and beyond lies the lake and still farther back the dark, volcanic range of the Butig Mountains on the southeast. To the southwest the striking peaks known locally as “Ganasi” appear. Usually, when the visitor first sees s Lake Lanao, if he has any curiosity at all, he seeks the most natural explanation in a volcanic region, namely, that it is a crater lake. There may be some resemblance to a crater rim on the south shore near Camp Vicars, but in other places there is no trace of it. I first was favorably inclined to the belief that it was a valley dammed by glacial wash and I found no trouble in likening the Keithley Escarpment to a terminal moraine. I was forced to abandon this hypothesis for reasons which will be stated in.a paper on the geology of this region which is to follow. My provisional conclusion with regard to Lake Lanao is that it occupies an old basin, partly tectonic, partly caused by erosion, between the mountains; this basin has been dammed by lava flows and other volcanic materials from the mountains adjacent to it. Subsequent weathering has given the aspect of a pseudo-glacial till to the material forming this obstruction. The explosion-crater theory has occurred to me and some attention will be paid to it in this con- nection in the later geologic discussion. Camp Keithley is situated partly on the brow of the escarpment of the same name and partly on the slope to the lake. The small village of Marahui lies on the lake shore on the west bank of the Agus. Here is the residence of the district governor and here too, the tribal court is reld. This village also has a native market, so that Marahui is the best place in the whole lake region to see the Moro people. Mr. Ickis joined the party in Marahui and we crossed Lake Lanao ina vinta with a large sail and awning made of some species of palm. It was almost nightfall when we ran into a small cove and landed, and in the darkness we began the ascent of 155 meters up the high bluff on which Camp Vicars is situated. This distance is between 4 and 5 kilo- meters. Three days later we set out for the Taraca River on the east side of the lake with an escort of twenty scouts, sixteen cargadores, and several guides. The country around Vicars is open and rolling and very similar to that south of Lake Lanao. Very little of the land is under cultivation. On the second day we finally descended from the high bluff we had been following and crossed an estuary, thus saving many miles of cir- cuitous travel. The low flats which border this side of the lake extend back for -several miles. The trail on the other side of this estuary les through paddies and swampy areas. Hyery morning during our march was clear and bright, but the afternoons without an exception were rainy. GEOLOGIC RECONNAISSANCE OF MINDANAO AND SULU. 491 The region through which we passed contains numerous Moro forts or walled towns termed “cottas” peculiar to the Lanao Lake district and to Sulu. The walls are several feet in thickness, made of earth, and protected by a dense hedge of bamboo growing at the top. A moat nearly always surrounds the cotta and a drawbridge of bamboo is provided. Bamboos, into the closed joints of which have been placed long, slender-barreled brass cannon, known as lantakas, are thrust through holes in the walls. The lantakas are imported from Singapore. We passed fifty or more of these cottas in our trip around this part of the lake. Some of them shelter only one or two houses, whereas others contain a score or more dwellings, mosques and other edifices, in fact an entire village. Each datw or sultan lives in his cotta with his family and retainers close about him, and there is constant petty warfare among the various chiefs. On the fourth day we reached the T'araca River and stopped at the house of a friendly datu. His cotta was, perhaps, the most elaborate we had seen; an elevation is shown by fig. 1. r Fie. 1. It was impracticable to ascend the Taraca River as far as the foothills to look for copper ore which had been reported from that point, because of the high water, the absence of trails along the bank, and the nature of the inhabitants. We did ascend for a distance of about 2 kilometers, but became almost hopelessly entangled in the ruins of old cottas, some recently destroyed by the Constabulary, others fallen into decay. This condition caused us to continue our march to Camp Keithley and from here we again crossed the lake to Camp Vicars, from which point we set out for Malabang. The first portions of the road lie across an open, almost treeless country which affords a splendid view of the Buldung Range which runs in a long, high, serrated line eastward from Malabang, Some six or eight extinct craters of different heights, arranged so as to resemble steps, are visible in this range. The road next enters heavy timber and at about one half the distance to Malabang, crosses the steel bridge over the Mataling River. The falls of Mataling are not so high as those of the Agus, but are scarcely less -492 SMITH. picturesque. The country rock is a basalt of more compact grain than at the other waterfall. The road runs into a very loose, black soil of sand and voleanic ash at a short distance beyond the bridge and continues in this formation to Malabang. ‘This deposit of ash, at some distant date, issued from the now extinct Buldung craters. The most notable feature at Malabang is the lne of cold springs issuing from the volcanic ash formation. We next proceeded by trail from Malabang to Parang. Apparently all the country rock at Parang consists of basalt with well-developed columnar structure about 500 meters above the pumping station at the military post. Just south of the town the basalt sheet suddenly ends and sedimentaries, including coal measures, appear. Carbonaceous shales and certain fossiliferous beds indicative of coal deposits are here found, although no coal has as yet been opened up. THE COTABATO REGION. The country from Parang to Cotabato is rolling and but scantily timbered. The country rock consists of shales and soft sandstones dip- ping southward, that is, toward Cotabato. There are several small lakes in this region which are noteworthy, because of the great profusion of large, pink lotus and the abundance of ducks. The difference between the topography in this region and that ar chal Malabang is due to the absence of the lava capping which becomes thin just to the south of Parang. Whereas the streams in the lava country have a cross section like the following figure (fig. 2), those in the Fie. 2. country to the southward beyond this sheet have more flaring sides to their valleys as is shown in fig. 3 Fie. 3. GEOLOGIC RECONNAISSANCE OF MINDANAO AND SULU. 493 FROM COTABATO TO DAVAO. The town of Cotabato is situated on the south bank of the north branch of the Rio Grande de Mindanao. It consists of a collection of low, white, Spanish houses fairly close together, with the usual native huts straggling about in the environs. Just south of and on the edge of the town is a limestone hill about 150 meters high from which a magnificent panorama (Plates XVI and XVII) can be obtained of the surrounding country. This hill is an outpost of a greater mass which is to be found to the south and which formerly was undoubt- edly still more extensive; it is very remarkable in that it stands out in the middle of the great delta, which forms all the river plain from the mouth of the river back and even beyond Fort Pikt. The most notable feature of the topography of this plain, beside the hill just mentioned, is the old terrace lines which swing along, but not always parallel to, either side of the river. These terraces are undoubtedly of marine origin, for close to Cotabato they are seen to be raised coral- reef shelves with the characteristic steep seaward slope of such formations. The evidence seems quite sufficient, to me at least, to suppose that. the sea once swept far up this intermontane region which is now so filled with sediment. Indeed, I am convinced that it one time extended through to the Gulf of Davao, for in the stretch between the Pulangui River and Davao there are very recent sediments and volcanics which have closed up the passage. Cotabato hill is composed of a cavernous limestone with a fair sprink- ling of fossils, corals, gasteropods, lamellibranchs, ete., all of compara- tively recent age, presumably Miocene, although no specific determinations have as yet been made. Plate XVIII shows the interior of a native salt-making establishment at the lower end of the delta. Sea water is sprayed over glowing embers, the salt is precipitated and afterwards washed off and run through the large filter shown in the background of the picture. This filter contains wood ashes and earth. At the Moro foundry near Cotabato, bolos, krises, and many metal boxes of brass and silver are fashioned. After some delay, we set out for Datw Piang’s place at Cuderangan, some 50 kilometers above Cotabato. Here we learned from Lieutenant Younglof, Philippine Scouts, of oil seeping from the river bank about half way between Reina Regente and Fort Pikit and also near Pilkit of a blue, plastic, oily clay which burns to a white color and is quite re- fractory. The Moros are said to come great distances to obtain this material. Reina Regente is on a hill of limestone similar to that of Cotabato. It is a monadnock. The underlying foundation is sandstone which will doubtless be found to be a good water carrier; it is very probable that good conditions for artesian wells can be obtained at almost any point of the valley. 78322 6 494 SMITH. Fort Pikit, which dates from the Spanish régime, like Reina Regente, surmounts a limestone monadnock, but the latter is much higher than the one on which the former fort is situated. This is the farthest port on the Rio Grande. From Fort Pikit we ascended to the end of navigation in the light- draft, stern-wheel steamboat which is used on the river. This point is some 50 kilometers beyond the fort, at the junction of the Kabacan with the Pulangui Rivers, the total ascent by steamboat being almost 200 knlometers. The first three days of our march were through mud; we were con- tinually forced to wade rivers, because we were following in the bed of the main stream, walking along the banks being out of the question, the first stop being at the junction of the Malabul and the Kabacan Rivers. We continued along the Malabul in a winding course, but making only 12 kilometers in a straight line in one day. A coarse, gritty sandstone and in places a typical conglomerate appear occasionally along the banks. On the third day after we left Pikit, and six days’ march from Davao, at an elevation of 365 meters, we reached the house of Datu Inkal, a Manobo chief. The geology in this region is not very prominent. ‘The trail generally leads through dense underbrush. All the streams are filled with large bowlders of extrusive rock, evidently from the Matutan Range just ahead. The latter is represented on the Jesuit map as a long, continuous and rather formidable Cordillera, but it is nothing of the kind and, except for Mounts Apo and Matutan, it is merely a broken - line of hills and quite low in several points. The journey for the next few days can best be given by extracts from the diary. December 16, 1907: Left Datu Inkal’s at 7 a. m. Continued through jungle and over rolling country to an elevation of 580 meters where the trail goes through the pass. Halted and made camp beside a small stream. Rainy weather and leeches made traveling very disagreeable. The feet of the cargadores were bleeding freely, but they did not seem to mind it. December 17, 1907: Broke camp at 7 a. m. Cloudy, elevation by barometer 472 meters. Crossed the Dalapnay River this morning. All the rocks for miles around this point appear to be similar, either fine-grained felsites, basalts and andesites or feldspar prophyries. Very little can be said geologically about this country at this time, as so little of it can really be seen. It apparently is extremely recent. Halted at noon at the Dalapnay River at a Manobo house and spent the afternoon of the 17th drying out our effects. Elevation at this point 412 meters. December 18, 1907: All of this day we are going downhill through dry woods, for the most part consisting of small trees and little or no underbrush. Occasional basalt and andesitic bowlders are seen. The difference between this side (eastern) and the western side of the range is almost entirely due to the fact that the prevailing winds, moisture-laden from the Sulu Sea, give up their moisture on the western side of the mountains and the winds blowing off the Pacific lose much of theirs on the seaside of the mountains east of Davao. The appearance of these eastern-slope forests is not greatly unlike that of those in the Temperate Zone. Mr McKinley Mt Penit Ape Me - GEOLOGIC RECONNAISSANCE OF MINDANAO AND SULU. 495 \ \ - . if —— = S Fra. 4. We halted for the night at 75 meters’ elevation at Sinauilan Creek, where soft, brownish-gray sandstone and conglomerate is exposed. December 19, 1907: Left camp at 6 a.m. Not much change either in topography or geology is apparent. Arrived at Digos, a small barrio on the coast of Davao Gulf, about 12 m., after a long walk over ground gently sloping to the beach. Here we spent the rest of the day and the night. December 20, 1907: We sent our cargadores and guard on to Davao by trail while we took the launch Bolinao which stopped off this point at noon. We arrived at Davao about 7 p. m. after stopping at two or three plantations on the way and went ashore the morning of the 21st. At Digos we obtained our first clear-cut view of Apo. The mountain stood out clearly and boldly, a sharp cone set to the south and back of an older truncated mass which had evidently blown off its head in some primordial paroxysm. On the southeastern side is a huge crevasse, from which puffs of a bluish-white vapor issue. Below 2,100 meters there is a dense jungle, a mass of green, but above this line the surface is all barren rock and apparently treeless, although when we ascended the mountain we found small bushes of blue- berries. The present high peak known as Apo did not pour the great mass of lava and rock over this entire region. The explosion crater was about 8 kilometers to the north- east. Fig. 4 shows a profile sketch of this mountain.® A wall back over the plain behind Davao is interest- ing. About 300 meters behind the town, or about 2 kalometers from the beach, unmistakable signs of old beach lines are found, marked by one distinct terrace at least 15 to 23 meters above the flat on which the town is located. All this territory is made up of alluvial wash from the hills. The bowlders are largely andesitic. Lieutenant Caffery left the party at Davao to return to Zamboanga. Without his assistance the reconnais- sance, up to the point where Davao was reached, would have been impossible. Daron, on the west side of the Gulf of Davao, was the starting point for the ascent of Mount Apo, the party consisting of Mr. Ickis, Mr. Goodman, who had just arrived from Manila, and Messrs. Carrigan and * A good picture of Mount Apo will be found in the article on Voleanoes and Seismic Centers, in the Census of the Phil- ippine Islands. (1903), 1, 201. 496 SMITH. McCall, the last named having the kindness to furnish a launch to take us from Davao to Daron, and our thanks are extended to him for the courtesy. We began the ascent of the first long, gradual slope toward Mount Apo on December 29. The trail first passes through hemp fields on the coastal plain and afterward it ascends gradually through a long, grassy slope which is strewn with occasional bowlders. The first stopping place was at the house of Tankalin, the chief of the Bagobos. In appearance he and his people are very much like the Manobos. ~. is 5 APE SAN A60STIN ~“. N, | s | be Siem a aneo® i. | Sarangani [ Balad re, SARANGAD! IS | it pels — anne, Sees =e — — —_—— as ———— o Tex? fer Longitide Kast of Greenwieh 125" 17" | fe 2 = : PLATE I- Smite: [PHIL. Journ. Sci., VoL. III, No. 6. ZAMBOANG BASILAN STRAIT SaaTH: RECONNAISSANCE OF MINDANAO AND SULU.] Ce: Sule : THE SULU ARCHIPELAGO Based on the latest Military x Geodetic Survey Maps and the Observations of Dr. N. M. Saleeby ° by 8 Pangutaran 1 aD Pandukon I. 2 J) *Veun, eS Pay NS r) ts, gq ea (. at ° SAHDAKAN HARBOR S ‘ Ploin Caralangan * SURVEYED BY FACING AND COMPASS eR 1. G. FERGUSON, MW. CLARK RIVER FROM U.S, COAST AND Scale |= 20.0 GEODETIC SURVEY CHART PS SS . SS COMTOUR iwrEAVAL 10 @ con, ourcrors Serreucen 90% A co rY pls waxsaiiy BOOK NOTICES. General Physics: An Elementary Text-book for Colleges. By Henry Crew, Ph. D. Pp., xiit522. Price, $2.75 net. New York: The Macmillan Company, 1908. The Elements of Physics. A College Text-Book. By Edward L. Nichols and William 8. Franklin. In three volumes: Vol. II, Electricity and Magnetism. Cloth. Pp., viii--522. Price, $1.60 net. New York: The Macmillan Com- pany, 1907. 78322——9 541 Jo Jeg la aN IU AV Page 41, Diospyrus canomoi should read Diospyros canomoi A. DC. Page 42, Antiaris toxicara should read Antiaris toxicaria Lesch. Page 44, Sunasia Amori Blanco should read Lunasia amara Blanco. Page 70, Table II, No. 22, 0.827 should read 0.927. Page 86, Under substances used for synthesis of ylang-ylang oil, add linaloél and geraniol. Page 146, Table VII, Calcium oxide should read 61.94% from bag; 62.04% from. can instead of 63.44% and 63.32%. Page 250, Under Filaria mosquito, arribalzaga should read Arribalzaga. Page 405, Voleanic fuel should read volcanic tuff. f 543 INDEX. [New generic and specific names Acacia catechu Willd., 534. Acanthopneuste borealis (Blasius), 280. Actitis hypoleucos (Linneus), 277. Adixoa tomentosa Schultze, 28. AGCAOILI, F. See GIBBS, H. D. Agglomerate (volcanic) in the Batanes Is- lands, 6, 7, 13. cutch made from, Agusan River, Mindanao, clay shale at the, | 508 ; earthquakes in the vicinity of the, 507; volume at Moncayo, 508; volume at Veruela, 505, 509. Alcedinide, 278. Alcedo bengalensis Brisson, 278. Alcohol from cassava, 93. Alkaloids in proprietary medicines, 412. Aluco longimembris (Jerdon), 283. Amaurornis pheenicura (Forster), 277. American Society specifications for cement testing, 167; 168; 171. Amorphophallus campanulatus Blume, 96. Analyses, chemical of clays from Botocan, 381; from Calamba, 384-387; from Los Banos, 383; from Majayjay, 380; Matiquio, 379; chemical of coal, methods used in tests of Philippine coal, 307, 309-10; of flue-gases, 307; of igne- ous rocks from Aroroy, 404; from Can- laon Volcano, 404; from Malaqui, Taal Volcano, 404; of tuff from Guadalupe, 404; from Majayjay, 404; proximate of | of flue- | gases, 315; proximate of Philippine coal, | coal ash and of clinker, 314; 312; ultimate of Philippine coal, 313. Analysis, of coal near Alcala, 537; of coal | from Tarragona, Davao, 502-503. Anatide, 277. Andesite at San Ramon, Zamboanga District, 482. Anthus gustavi Swinhoe, 281. Anthus rufulus Vieillot, 281. Antiarin, 42. Antiaris toxicaria Lesch., 42. Apo, Mount, 477; ascent of, height determination of, 497. Apparatus used for coal tests, 303-310: Boiler, factors by, 348-349, 352; used in steaming test of Philippine and other coals, 304; Chimney, 304; Fire box, 304; elongated, 342-345, 353; the effect of water in the air on the economy of, 350-351, 354; Grate, 302-304, 311, 342-345, 353; loss through the, 302, 345. from | 309; | 496-498 ; | influencing absorption | are printed in heavy-faced type.] | Arbelide, 29. | Arctiide, 30. | Ardeidz, 277. | arenata (Squamicapilla), 30. Arrow poisons of the Philippines, 41. Arrowroot (Maranta arundinacea Linn.), 96. | Artamide, 280. Artamus leucorynchus (Linnzus), 280. | Aspidomorpha miliaris Fabr., 264. | Ash, analyses of, of Philippine coals and others offered for sale on the Manila | market, 314; color of coal, 302, 312, | 341; from Philippine and other coals, 310, 314, 354; the influence on the efficiency of the combustible, 341; the relation of its content to the external appearance of Philippine coals, 91. ; Australian coal test, Lichzow Valley, 311— | 316, 319, 326-327; Westwaldsend, 311— | 316, 319, 323-325. | BACON, RAYMOND F., Editorial: Starch | production in the Philippine Islands, 93; Editorial: Philippine arrow poisons, 41; Philippine terpenes and essential oils, I, 49; Philippine terpenes and essential oils, II, ylang-ylang oil, 65. Baffle wall, influence of, on the combustion of bituminous coals, 342-344, 353. | Banglag Creek, Davao, fossiliferous clay at the, 507; sandstone at the, 507. BANKS, CHARLES §., Biology of Philippine Culicide, 235. | Banksinella luteolateralis luteolateralis Theob., 254. | Basalt at Basilan, 487; at Jol6, 485. Bashi Channel, 13. | Basilan Island, 487. basipuncta (Crusiseta), 33. ' Batanes Islands, climate of, 4; geographical description of, including Batan, 7; Des- quey, 12; Ibujos, 12; Inem, 11; Isbayat, 12; Sabtan, 6; the Siayanes, 11; geologic problems of: alignment of volcanoes, 14; correlation with Formosa and the Babu- yanes, 15; faulting, 14; origin of agglom- erate, 13; history of, 3; location of, 2; people of, 3; physiography of, 16; sub- marine configuration of, 12. Batan Island (Batanes), 7, 16, 19. | BEAN, ROBERT BENNETT, A theory of heredity to explain the types of the white race, 215; The Benguet Igorots. A so- matologic study of the live folk of Ben- guet and Lepanto-Bontoc, 413. 545 Theob.; Culex 546 Beetles, typhoons, and coconuts, 533. Benzoic acid in foods and drugs, 101. Benzoin, 59. Benzyl benzoate, 85; benzyl formate, benzyl-methyl ether, 85; benzyl salicylate, 85; benzyl valerianate, 85. Betts’ mine, Batan Island, coal test, 316, 321, 337-338. Biology of Philippine Culicide, 235; Banks- inella luteolateralis Theob., 254; Des- yoidya joloensis Ludlow, 240; oviposition, 240; habits of the larve, 241; Hulecoe- tomyia pseudoteniata Giles, 249; habits of the adult, 250; habits of the larve, 249; Mansonia annulifera Theob., 255; Mansonia uniformis Theob., Stego- myia persistans Banks, 243; habits of the adult, 245; Stegomyia samarensis Ludlow, 246 ; oviposition, 247 ; habits of the adult, 248; the Filaria mosquito, 250; habits of the adult, 253; Worcesteria grata Banks, 235; habits of the adult, 239; habits of the larve, 238. Birds, notes on a collection of, from Siquijor, 275. 25D ; Boalon, limestone at, 482; schist at. 452. Boiler, factors influencing 348-349, 352; used in steaming test of Philippine and other coals, 304. Boiler pressure, Maintained during steaming tests of Philippine and other coals, 311; the effect of, on efficiency, 349, 353. Bongao Island, 486. Book notices : Crew, Henry, General physics: An ele- mentary text-book for colleges, 541; | Nichols, Edward L., and Franklin, William S., The elements of physics : A college text-book, 541. Borneo coal test, Labuan, 330-331, 311-316, 320. Boron in butters and hams, 45; in foods and drugs, 101. Botel Tobago, 15. Botocan clays, 381; chemical analysis of, 381; physical properties of, 382. Briquettes, machine used for breaking sand, 167; moist-air closet for the storing of, 153; tensile strength of neat and sand, 160. browni (Deilemera), 31. Bruguiera eriopetala W. & A., B. gymmor- rhiza Lamk., bark extract made from, 535. Bubonide, 278. Bubuleus coromandus (Boddaert), 277. Bud Dajo, Jolo Island, 485. Butter, boron in, 45; detection of coconut oil in, 371; examination of, 45. Cacatua hematuropygia (P. L. S. Miiller), 278. Cacatuide, 278. Cacomantis merulinus (Scopoli), 278. Calamba, Pajo Canon and Point Alipasio clays, 384; chemical analysis and physical properties of, 384-388. 85; | 311— | absorptions by, | INDEX. | Calenas nicobarica (Linneus), 277. | Galorifie value, determination, 310; estima- tion, 309; of Philippine coals and others offered for sale on the Manila market, 313, 344, 354. | Camoting cahoy, 93. Campophagide, 279. Canangium odoratum Baill., 65. Cassava plant (Manihot utilissima Pohl), 93. Cassida (Odontionycha) picifrons Weise, 259, 266. Cassidide, description of new, of the Phil- ippine Islands, 259; life histories of some Philippine, 261. Catechol in phenol, 364. Cement, American Society specifications for testing of, 167, 168, 170, 171; climatic influences on the testing of, 176-178, 180; effects of aération on, 141; influence of temperature on time of setting of, 151; specific gravity and loss on ignition in the testing of, 171; tamper for the testing of, 161-163; testing of Portland, 137; the disposal of caked, 150; time of setting of, 155; United States Army specifications for cement testing, 155, 161, 164, 168, 169, 171; use of volcanic tuff as a mate- rial for the manufacture of, 404-406. Centropus viridis (Scopoli), 279. Cephalophoneus nasutus (Scopoli), 280. Ceriops roxburghiana Arn., bark. extract made from, 535. | Ceryx macgregori Schultze, | Chaleophaps indica (Linnzus), 276. | Charadriide, 277. Charadrius fulvus (Gmelin), 277. Cheese, examination of, 45. Chimney gases, 345-348 ; analysis of, 307; analyses of, during steaming tests of Phil- ippine coals and others offered for sale | on the Manila market, 315. Chinese medicines, 411. Cicindelide, a uew species of Philippine, | Prothyma schultzei Horn, 273. 29 Cinnyris henkei Meyer, 283. Ginnyris jugularis (Linneus), 280. Cinnyris sperata (Linnzus), 250. Cisticola exilis (Vigors and Horsfield), 280. Clay, at the Agusan River, 509; fossilifer- ous, at the Banglag Creek, Dayao, 507 ; at the Mauntoc River, Davao, 506. | Clays, Botocan, 381; Laguna, 377; Majay- | jay, 380; Matiquio, 378; Mount Maqui- | ling region (Calamba, Los Bafios), 382. Climate, influence of, on* cement testing, 176-178, 180; of Mindanao, 479; of the Batanes Islands, 4. Clinker, 341-342, 354; analyses of, of Phil- ippine coals and others offered for sale on the Manila market, 314; in the refuse from Philippine and other coals, 302, 314, 341, 354. INDEX. DAT Coal, analysis of, from near Alcala, 537; | Conglomerate at the Banglag Creek, 507; analysis of, from Tarragona, Davao, 503; at Lucatan, Dayao, 502; at Mount Mayo, Australian test of: Lichzow Valley, 311— 316, 319, 326-327; Westwaldsend, 311— 316, 319, 323-325; Betts’ mine, Batan Island, test of, 311-316, 321, 337, 338; Borneo test of, Labuan, 330-331, 311—316, 320; Comansi mine, near Danao, Cebu, test of, 311-316, 321, 339-340 ; impurities in, 341, 353; Japan test of: Yoshinotani (Karatsu), Kiushu Island, 311-316, 319, 328; Yubari (Hokkaido Province), 311— 316, 320, 329; loss due to incomplete combustion of, 345-348; Military Reser- vation, Batan Island, test of, 311-316, 320-321, 332-336; outcrops of, at Bag- gao, 535; at Nasiping, 536; Polillo, test of, 311-316, 321; rapid expulsion of the volatile matter of Philippine, 352; Sibu- guey, Zamboanga District, 483; tendency of Philippine to fall to pieces, 354; the physical condition of, 311, 350, 354; the relationship between the external appear- ance and the ash content of Philippine, Gals Coals, calorific value of Philippine and others offered for sale on the Manila market, 313, 344, 354; clinker from Phil- ippine and others, 302, 314, 341, 354; color of smoke of Philippine and others, 310, 315, 354; distribution of the heating value of the combustible of, 319-321; equivalent evaporation of water from and at 100° C. of Philippine and others offered for sale on the Manila market, 316; importance of the size of, for fuel, 354; influence of baffle wall on the combustion of, 342-344, 353; kind of furnace suc- cessfully to burn Philippine, observation in detail of the tests of, 323— 340; Philippine and others arranged in order of decrease in ash content, 91; Philippine as fuel, 301; proximate anal- ysis of Philippine and others, 312; rate of evaporation of water by Philippine, 348; specific gravity of, 312; steaming tests of, 311—316 ; the relation of the ash content to the appearance of Philippine, | 91; ultimate analysis of Philippine and other, 313; United States Army transport | tests of Philippine, 302. Coconut, changes in the composition of, | while sprouting, 119; experiments on | enzymes in the, 111; notes on the sprout- ing, 111. Coconut oil, notes on the sprouting coconut, on copra, 111; on the detection and deter- | mination of, 371; purification of, 45; the production of free acid in commercial, on long standing, 126. Coconuts, typhoons, and beetles, 533. COLE, FAY COOPER, The Tinggian, 197. Colloecalia troglodytes Gray, 278. Comansi mine, near Danao, Cebu, coal test of, 311-316, 321, 339, 340. 502. Constants of first-grade ylang-ylang oils, 70. COPELAND, EDWIN B., Editorial: Ty- phoons, coconuts, and beetles, 533. Copra, notes on the sprouting coconut, on coconut oil and on, 111; the action of 92 microorganisms in pure culture on, Copsychus mindanensis (Gmelin), 279. Coral in the Batanes Islands, 21. Corone philippina (Bonaparte), 281. Corvide, 281. COX, ALVIN J., Editorial: The effect of Litsea chinensis on the hardening of lime mortar, 409; Editorial: The relationship between the external appearance and the ash content of Philippine coal, 91; La- guna clays, 377; Philippine coals as fuel, 301; Voleanic tuff as a construction and a cement material, 391. Crusiseta basipuncta Schultze, 33. Cuculide, 278. Culex aestuans Wied., 23. C. anxifer Coquerel (Bigot), C. fatigans Wied., C. macleayi Skuse, C. pallipes Meigen, C. pungens Wied., Heteronycha dolosa Arribalzaga, 250. Culicide, biology of Philippine, 235. Cuteh, 534. Cyecas circinalis Linn., 96. Cyornis philippinensis Sharpe, 279. Deilemera browni Schultze, 31. delicata (Psecadia), 36. Al para menthene, 52. Dendrocygna arcuata (Horsfield), 277. Desquey Island (Batanes), 12, 18. 342 344- | Desvoidya joloensis Ludlow ; Desvoidea fusca joloensis Ludlow; Desyoidya fusca jolo- ensis Banks, 240. Diabase at Pujada Peninsula, 501. Diceide, 280. Diceum besti Steere, 280. Diceum pygmeum (Kittlitz), 280. diffusihelvola (Euchromia), 29. Dihydro-limonene, 52. Dioscorea sp., 96. Diospyros canomoi A. DC., 41. Distribution of the heating value of the combustible in tests of Philippine coals and others offered for sale on the Manila market, 319-321. Dita bark, 44. Draft, 304, 345; average force of, in steam- ing tests of Philippine coals, 312. Dysentery cure, 412. Earthquakes in the vicinity of the Agusan River, Mindanao, 507. elegantissima (Euchromia), 29. Elymnias palmifolia Schultze, 2 ENRIQUEZ, PIO VALENCIA. DOZA, MARIA P. Enzymes, 411. See MEN- experiments on, in the coconut, 548 Euchromia elegantissima Wllgr., 29. Eudynamis mindanensis (Linneus), 278. Evaporation of water, equivalent from and at 100° of Philippine coals and others offered for sale on the Manila market, 316, Excalfactoria lineata (Scopoli), 276. Falco ernesti Sharpe, 277. Falconide, 277. Faulting in the Batanes Islands, 14. FERGUSON, HENRY G., Editorial: Coal in the Cagayan Valley, 535; Contributions to the physiography of the Philippine Is- lands: II, The Batanes Islands, 1. ferrugineum (Tribolium), 299. Filaria mosquito: Culex estuans Wied., C. anxifer Coquerel (Bigot), C. fatigans Wied., C. macleayi Skuse, C. pallipes Meigen, C. pungens Wied., Heteronycha dolosa Arribalzaga, 250. Fire box, 304; elongated, 342-345, the effect of water in the air on economy of, 350-351, 354. Fishes, important commercial of the Philip- 353 ; the pine Islands: Anchovies, 513; herrings, 514; mackerels, 515; milk, 519; mud, 516; mullets, 517; pampanos, 517; sea basses, 517; silversides, 514; snappers, 516; index to Philippine, 528. Fishing, methods of, 521; .ocation of, banks, 528. Fish ponds, 519. Formic acid in ylang-ylang oils, 78. FOXWORTHY, F. W., Editorial: 534. Cutch, Fuel, importance of the size of coal used as, | 354; Philippine coals used as, 301. Gallus gallus (Linneus), 276. Geology of the Batanes Islands, 1. Geometride, 34. Geraniol benzoate, 86. Geraniol-methyl ether, 86. Gibbium scotias Fabr., notes on the abun- | dant appearance of, in the Philippine Is- | lands, 299. GIBBS, H. D., Editorial: Food and drug inspection, 44; Methyl salicylate, I, The separation of salicylic acid from methyl salicylate and the hydrolysis of the ester, | II, Solubility in| 101; Methyl salicylate, water at 30°, 357; On the detection and determination of coconut oil, 371; Pro- prietary medicines in the orient, 411; The compounds which cause the red color’ in phenol, 361. GOODMAN, MAURICE, A _ reconnaissance from Davao, Mindanao, over the divide of | the Sahug River to Butuan, including a survey from Davao to Mati. Narrative of the expedition, 501. Grate, 302-304; 311, 342-345, 353; through the, 302, 345. Grignard reaction on terpenes, 50. Haleyon chloris (Boddaert), 278. Halcyon gularis (Kuhl), 278. Haliastur intermedius Gurney, 277. Ham, boron in, 45; examination of, 45. loss INDEX. Heat balance of the heating value of the combustible in tests of Philippine coals and others offered for sale on the Manila market, 319-321. Hemichelidon griseisticta (Swinhoe), 279. Heredity to explain the types of the white race in North America, a theory of, 215; arrangement of data of, 215; bibliography of, 229; classification of types, 215; femi- nine types, 218; types, 219. Hirundinide, 279. Hirundo javanica Sparrmann, 279. Horlick’s malted milk, the composition of, 87. HORN, WALTHER, Prothyma schultzei, a new species of Philippine Cicindelide, iis Hulecoetomyia pseudoteniata Giles; Hule- coeteomyia pseudoteniata Theob.; Stego- myia pseudoteniata Giles, 249. Hydrolysis of methyl salicylate, 101. Hydroxyl groups, method of estimation ac- cording to Zerewitinoff, 83. Hyloterpe apoensis Mearns, 280. Hypotenidia torquata (Linneus), 277. Hypothymis occipitalis (Vigors), 279. Ibujos Island (Batanes), 12, 18. Igorots of Benguet, 413; descriptive char- acters of, 441; Malay vs., 440; methods employed in examination of, 415; local- ity of, 4138-414; physiognomy of, 436; proportion of the body parts of, 423; selected types of, 454; somatologic race types of, 451; stature of, 417; supple- mentary theory of heredity of, 460. Inem Island (Batanes), 11. integra (Pericallia), 30. intextilia (Remigia), 32. Iole siquijorensis Steere, 279. Iraya Volcano (Batanes), 9. Isbayat Island (Batanes), 12. : Japan coal test, Yoshinotani (Karatsu) Kiushu Island, 311-316, 319, 328; Yu- bari (Hokkaido Province), 311-316, 320, 329. Jolo, 484—486. juvencus (Sirex), 299. Kaolin, chemical analyses and physical prop- erties of American, 378; theoretical com- position of pure, 377. Keithley Escarpment, Lanao District, 489. Kinabuungan, Mount, Davao, 506. Lacoptera philippinensis Blanch., 268. Lalage niger (Forster), 279. Lamprocorax panayensis (Scopoli), 281. Lanao Lake, Mindanao, 490. Laniide, 280. Laride, 277. Lepidoptera, new and little-known, of the Philippine Islands, 27. Lime mortar, effect of Litsea chinensis Lam. on the hardening of, 409. Limestone, at Boalon, Zamboanga District, 482; at Cotabato, 493; at Lucatan, Da- vao, 502; at the Sahug River, Davao, 505; of the Batanes Islands, 6, 10, 12. Limonene hydrochloride, 51. INDEX. Linalool-methyl ether, 86. Linao, Lake, Agusan, 509. Litsea chinensis Lam., effect of, hardening of lime mortar, 409. Lophopetalum toxicum Loher, 44. Loriculus siquijorensis Steere, 278. Los Banos clays, 382; chemical analyses of, 383; physical properties of, 383-384. lucidicollis (Prothyma), 273. Lunasia amara Blanco, 44. Maasam River, Agusan, stone at, 510. Mabudis Island (Batanes), 11. macgregori (Ceryx), 29. Magnesium action on terpene hydrohalides, 49, Majayjay clays, 380; chemical analysis of, 380; physical properties of, 381. Malindang, Mount, Lanao District, 488. Malted milk, the composition of Horlick’s, 87. Manihot utilissima Pohl, 93. Manila elemi, 49. Manioc, 93. Mansonia annulifera Theob., Panoplites an- nulifera Theob., 255. Mansonia uniformis Theob., Panoplites uni- on the argillaceous sand- | 549 ; Modeling, an improved method of, especially | i | Mount Maquiling region clays. adapted for the central nervous system, 293. Mortar, method to determine the setting time of, 158. Motacilla melanope Pallas, 281. Motacillide, 281. See Calam- ba and Los Baios. Muscadivores chalybura (Bonaparte), 276. Muscicapide, 279. MUSGRAVE, W. GEORGE F. Myristicivora bicolor (Scopoli), 276. Narcotic drug law, 412. Nectariniide, 280. Ninox philippensis Bonaparte, 278. Noctuide, 31. Nycticorax manillensis Vigors, 277. Nymphalide, 27. Oil, detection and determination of coconut, 371. E. See RICHMOND, | opala (Polydesma), 31. Oriolide, 281. | Oriolus chinensis Linneus, 281. formis Theob., Mansonia africanus Theob., | Mansonia australianis Giles, 255. Maranta arundinacea Linn., 96. Marine erosion in the Batanes Islands, 19. Matiquio clays, 378; chemical analysis of, 379; physical properties of, 379. matuta (Pyrausta), 36. Mayo, Mount, Davao, conglomerate at, 502. May-sanga Island (Batanes), 11. McGREGOR, RICHARD C., Notes on a col- lection of birds from Siquijor, Philippine Islands, 275; Philippine ornithological literature, I, 285; Some necessary changes in the names of Philippine birds, 283. Megalurus tweeddalei McGregor, 283. MENDOZA, MARIA P., RAMIREZ, MA- NUEL, & ENRIQUEZ, PIO VALENCIA., An improved method of modeling espe- Ornithology, literature of Philippine, 285. Oryctes rhinoceros Linn., as a menace to coconuts, 533. | Osmotreron axillaris (Bonaparte), 276. cially adapted for the central nervous | system. 293. Meropide, 278. Merops philippinus Linneus, 278. Methyl salicylate, analytical determination of, 357; hydrolysis of, 101; in pharma- copoeial preparations, 101; in root beer, 101; in soda water flavors, 101; solubil- ity of, 357. Metriona trivittata Fabr., 267. miliaris (Aspidomorpha), 264. “Milionia pretiosa Schultze, 34. Military Reservation, Batan Island, test, 311-316, 320-321, 332-336, Milk, the composition of Horlick’s malted, 87. MILLER, MERTON L., Editorial : cent of Mount Pulog, 99. Preparation of brain models, coal The as- geographic description of, 476; people of, 478; rainfall at, 479; temperature: of, 480. Osmotreron vernans (Linnezus), 276. Otomela lucionensis (Linnzus), 280. Oxygen equivalent of oils, 371. Pachyrhizus bulbosus Kurz (P. Rich.), 96. palmifolia (Elymnias), 27. People of the Batanes Islands, 3, Pericallia integra Wlk., 30. Peridinium in Manila Bay, 187. Peristeride, 276. Petrophila manilla Boddaert, 279. Phapitreron albifrons McGregor, 276. Pharmacopoeial preparations, salicylic acid and methyl salicylate in, 101. Phasianide, 276. Phenol, catechol in, 364; cause of red color in, 361; oxidation of, 363; phenoquinone in, 364; quinone in, 364. Phenoquinone in phenol, 364. Philippine birds, some necessary changes in the names of, 283. angulatus 23. Philippine coals. See Coals. Philippine Islands coal test. See Coal. philippinensis (Lacoptera), 268. Physical properties of clays from Botocan, 382; from Calamba, 385-388; from Los Bafios, 383-384; from Majayjay, 381; from Matiquio, 379; of coal, 91, 311, 350, 354. Physiography of the Batanes Islands, 1, 16. Picifrons (Cassida), 259, 266. Pikit, Fort, 494. Pitta atricapilla Lesson, 279. Pittide, 279. | Polillo coal test, 311-316, 321. Mindanao, climate of, 479; geology of, 473; | Polydesma opala Pagents., 31. Pratincola caprata (Linneus), 279. pretiosa (Milionia), 34. | Prioptera schultzei Weise, 259, 263. 5DO Prioptera sinuata Oliy., 261. Proprietary medicines in the orient, 411. Prothyma lucidicollis Chd., 273. Psecadia delicata Schultze, 36. Psittacide, 278. Pulog, the ascent of Mount, 99. Puso-puso (Litsea chinensis Lam.), 409. Pycnonotide, 279. Pyralide, 34. Pyrausta matuta Schultze, Pyrausta vastatrix Schultze, 35. Quinone in phenol, 364. . Radiation of heat, 319-321, 350. Rainfall at Mindanao, 479; in the Batanes Islands, 4. Rallide, 277. RAMIREZ, MANUEL. RIA P. 36. INDEX. SCHULTZE, W., Editorials: Notes on the abundant appearance of Gibbium scotias Fabr., in the Philippine Islands, 299; Notes on the appearance of Sirex juven- cus Linn., in Manila, P. I., 299; Life histories of some Philippine Cassidide, 261; New and little-known Lepidoptera of the Philippine Islands, 27. | Scirpophaga virginia Schultze, 34. See MENDOZA, MA- | REIBLING, W. C., & SALINGER, L. A., Portland cement testing, 137. Remigia intextilia Schultze, 32. Rhipidura nigritorquis Vigors, 279. Rhizophora mucronata Lamk., R. conjugata Lamk., bark extract made from, 535. Rhynchophorus ferrugineus JFfabr. as a_ menace to coconuts, 533. RICHMOND, GEORGE F., Editorial: Puri- | fication of coconut oil, 45. RICHMOND, GEORGE F., & MUSGRAVE, | W. E., The composition of MHorlick’s malted milk, 87. Root beer, salicylic acid and methyl sali- cylate in, 101. Sabtan Island (Batanes), 6, 16, 17. Safrol (isosafrol) in ylang-ylang oil, 78. Sago, 96. Sahug River, Davao, 504—506; sandstone at the, 504. Salicylic acid, determination and separation in foods and drugs of methyl salicylate and, 101; in flavors for soda water, 101; in foods and drugs, 101; in pharmaco- poeial preparations, 101; root beer, 101; in sarsaparilla, 101. SALINGER, L. A. See REIBLING, W. C. Sandstone, at the Banglag Creek, Davao, 507; at the Maasam River, Agusan, 510; at Mount Kinabuungan, Davao, 506; at the Sahug River, Davao, 504—506 ; Cotabato region, 492. San Ramon, Zamboanga bowlders andesite fragments of schist at, 482. Sarcops melanonotus Grant, 281. Sarsaparilla, salicylic acid in, 101. Seale, the effect of in the transmission of heat through boilers, 349. Schist, at Boalon, Zamboanga District, 482 ; chloritic, at Pujada Peninsula, 502. schultzei (Prioptera), 259, 263. schultzei (Prothyma), 273. in in the District, containing 482; of large scotias (Gibbium), 299. Sea horses in medicines, 411. SEALE, ALVIN, The fishery resources of the Philippine Islands. Part I, Commer- cial fishes, 513. Sesiide, 28. Siasi Island, 487. Siayanes Islands (Batanes), 11. Sibuguey, Zamboanga District, coal at, 483. Sincamas (Pachyrhizus bulbosus Kurz), 96. sinuata (Prioptera), 261. Sirex juvencus Linn., notes on the appear- ance of, in Manila, 299. SMITH, WARREN D., A_ geologic recon- naissance of the Island of Mindanao and the Sulu Archipelago. I, Narrative of the expedition, 473; Editorial: An account of a human sacrifice held by the Bagobos, District of Davao, Mindanao, P. I., 188. Smoke, color of, as observed when Philip- pine coals and others offered for sale on the Manila market are fired, 310, 315, 354. See also Chimney gases. Soda water flavors, salicylic acid and methyl salicylate in, 101. Solubility of methyl salicylate, 357. Specific gravity of coals, 312; loss on igni- tion in cement testing, 171. splendida (Vitessa), 35. Squamicapilla arenata Schultze, 30. Squatarola squatarola (Linneus), 277. Starch production in the Philippine Islands, 93. Steam, calorimeter, 305-306; quality of that produced during tests of Philippine and other coals, 306-307. Stegomyia persistans Banks, S. fasciata per- sistans Banks, 243. Stegomyia samarensis Ludlow, 246. Sterna boreotis (Bangs), 277. Stock food, 93. Strength, crushing, of andesitic tuff from near Manila, 394; tensile, of stone being used in Manila, 395; tensile, of Majayjay stone, 395; transverse, 396; transverse, of Wisconsin building stone, 397. Streptopelia dussumieri (Temminck), 276. Strophanthus cumingii A. DC., 44. Sturnide, 281. Submarine configuration of the Batanes Is- lands, 12. Sulu Archipelago, 484. Sylviidz, 280. Syntomide, 29. Tacca pinnatifida Forst., 96- Tagum River, Davao, 503-504. Tamper for cement testing, 161; illustra- tions of, 162-163. INDEX. dd1 Tanygnathus lucionensis (Linneus), 278. Taraca River, Lanao District, 491. Tarragona, Davao, coal from, 502-503. Tawi-Tawi Island, 486. Temperature, influence of, on time of setting cement, 151; of Mindanao, 480; of the Batanes Islands, 5. Terpenes and essential oils of the Philip- pines, 49; Terpenes and essential oils of the Philippines, ylang-ylang oil, 65; ana- lysis of ylang-ylang oil, 69; composition of ylang-ylang oil, 78; constants of first- grade ylang-ylang oils, 70; general trade conditions, 66; hydroxyl groups, method | of estimation, Zerewitinoff, 83: methods of distillation, 68; synthesis of ylang- ylang oil, 86; the adulteration of ylang- ylang oil, 74. Tests, American Society specifications for cement, 167, 168, 170, 171; climatic in- fluences on cement, 176-178, 180; for tomentosa (Adixoa), 28. Treronide, 276. Tribolium ferrugineum Fabr., notes on the abundant appearance of, in the Philippine Islands, 299. | trivittata (Metriona), 267. soundness, 158; specific grayity and loss on ignition in cement, 171; Portland cement, 137; tamper for cement, 161— 163; United States Army specifications for cement, 155, 161, 164, 168, 169, 171. Tests (steam of coal), average barometer reading during, 311; average steam pres- sure maintained during, 311; description of apparatus and methods employed, 303— 310; detailed observations of, on coals from Australia, 325-327; from Borneo, 330-331; from Japan, 328—329 ; from the Philippine Islands, 332-340; equivalent evaporation of water from and at 100° C. | of Philippine coals and others offered for sale on the Manila market, 316; flue gas, 307-309 ; kind of grate used in, 302-303, | 311, 342-345, 353; lowering of the value | of, by moisture in the air, when made in the Tropics, 350-351, 354; methods of | analysis employed, 307, 309, 310; neces- | sity for uniform conditions in making, 303; number of, necessary for valuation, © 303; Philippine coals and others offered for sale on the Manila market, 311—318. Test with dried and undried cement taken from paper bags, 144. Tetrahbydro-limonene, 54. Tidal scour in the Batanes Islands, 20. Tineide, 36. Tinggian, a Philippine people, 197; births and marriages of, 206; funerals of, 210; geographical distribution and migration of, 198; government of, 203; physique, — dress, and customs of, 199; religion of, | 204. Turdide, 279. Typhoons, at Mindanao, 481; beetles, co- conuts, and, 533. United States Army specifications for cement testing, 155; 161; 164; 168; 169; 171. vastatrix (Pyrausta), 35. Veruela, Agusan River, 505; 509. virginia (Scirpophaga), 34. Vitessa splendida Schultze, 35. Voleanic tuff as a construction and a cement material, 391; cement, 404; chemical examination of, 404, cubic weight of, 399; curve from which the time required to dry a cube of any size may be directly read, 403; curve showing the rate of drying in the air, 401; durability of, 397; microscopical examination of, 393; porosity of, 398; specific gravity of, 397 ; table of crushing strength of, from several quarries near Manila, 394; of tensile strength of Majayjay, 395; tensile strength of, being used in Manila, 395; the rate at which the drying takes place in the air, 400; of transverse strength, 396; of transverse strength of Wisconsin building stone, 397; temperature changes, 398. Volcanoes, alignment of, 14. Vulcanism, effects of, in the Batanes Is- lands, 22. WALKER, HERBERT §&., Notes on the sprouting coconut, on copra, and on ¢co- conut oil, 111. Water, equivalent evaporation from and at 100° of Philippine coals and others offered for sale on the Manila market, 316; in air, the effect on the economy of furnaces, 350-351, 354; in steam produced during tests of Philippine and other coals, 306— 307. WEISE, J., Description of new Cassidide of the Philippine Islands, 259. Worcesteria grata Banks, 235. Y’Ami Island (Batanes), 11. Ylang-ylang oil, 65. ; Zamboanga, Mindanao, 481. | Zosteropide, 280. Zosterops boholensis McGregor, 283. Zosterops siquijorensis Bourns and Worces- ter, 280. A Nay iene ti By coin MUSGRAVE: STREPTOTURICOSIS. Pui. Journ. Scr, Vor. II, No. 6. 1 ; Fig. 2. Fig. 1. PLATE I. ¢ = ae “a us ive ; ao oD re Reels See! MUSGRAVE: STREPTOTHRICOSIS.] (Putt. Journ. Scr, Vor. I, No. 6. Fig. 3, PLATE ll. MUSGRAVE: STREPLOTHRICOsIS.] [Putt Journ. Scr, Vou. Hl, No. 6. . Fig. 5. PLATE Ill. MUusGRAVE: SPREPTOTHRICOSIS.] [Purt. Journ. Scr, Vou. IL, No. 6. Fig. 8. Fig. 7. PLATE IV. MUSGRAVE: STREPTOTHRICOSIS.| (Put. Journ. Scr, Vou, I, No. 6. Fig. 10. Fig. 9. PLATE V. MUSGRAVE: StREPTOTHRICOSIS.] (Pum. Journ. Scr, Vou. Il, No. 6. Fig. 12. Fig. Il. PLATE VI. recta arate ftir ohren em me? S179 WiO ad Ose et ee ee a enn np ehiton at thove einen aleneine nee ytreket EPA bIANviEt gba! am Roar ererecy ras Neeru encanta frie Robetihiower Caieh ht st ihie SMITHSONIAN INSTITUTION LIBRARIES . elecgte ta 7 Roueecnsyet 88 01307 6641 see fie degena ss naganer eh § weed WV nest steme vacate mebmen tt me eh hal her mrenicrees Le deb re bot ie 18 A Fob ahe MINA Ib oe Mae GRR MM ALE oh Pear Ce! reer Tyrvcrenin A AOUIE De hseetes ey Dieter Ret tat Abe bi Peaciiibek sie twat peiiih ont Ay MBERLALs Laaaas inom inde be it Roerti oe acek sat Chk 4 Alban SOIR criaaet Woraeg ont fhe Coser tev arta tite ere i Mmlwias anys ange hee raek Fhoy Cecieeint Smith bate ean ens Sora eer Vey ipieietwiwtye pete ACEABACH doh Siw NTL TW wigryth Sy wroreeree! Ryo trite) 4 aie aR OR mh Ab Hebe y abe oh DHFR Usuerbey lier wrecue mii MEOIN wel ule veswih wf oa AiAlaisey ibibibobebrhsweatbinen er hewe Web athsashea chibcet yee thw fay ho bem wth aar yg Wineee chen muchey erie arcs te lerwaccarnen ied sMUMeMTa sflaqnibe oy SUA ASGER Winepiiepissnrererecalbiaruce Piedra armersombare, | Woh ih ih mie ates AMUN Ackiy hen seetalarh oath renters jor emiie brismewiare ere tee nt